From Wagner to Taft-Hartley, Revisited

Anthony Michael Daniel

Submitted in partial fulfillment of therequirements for the degree of

Doctor of Philosophyin the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY2017

c�2017Anthony Michael Daniel

All rights reserved

ABSTRACT

From Wagner to Taft-Hartley, Revisited

Anthony Michael Daniel

This dissertation is guided by two questions: why was the right to collective,

abrupt cessation of work and production sharply truncated at the height of labor’s

power during the New Deal; and, further, what does this mean for our polity? The

extant laws governing workplace protest have gone unchanged since the 1947 Taft-

Hartley amendments to the 1935 Wagner Act. This dissertation revisits the Wagner

to Taft-Hartley period - 1935 to 1947 - to identify commonalities in the politics sur-

rounding state adoption of Taft-Hartley precursor laws, particularly those pertaining

to the right to strike.

In the first chapter, I outline the motivation for revisiting the New Deal period

and the specific importance of union militancy for realizing working-class program-

matic aspirations on the shop floor and beyond. I also outline the key empirical puzzle

of my dissertation: why did some states adopt anti-strike laws while others did not?

In the second chapter, I analyze the volume and content of the newspaper cov-

erage of labor militance during the New Deal period through Reconversion with a

particular focus on landmark historical episodes. In general, I find that coverage of

strikes mirrored their incidence. However, the content of the coverage of strikes tended

to carry more negative source-messages than positive. Further, the coverage tended to

decontextualize the class conflict wrought by strikes. These content analyses suggest

that the apparent lack of class consciousness on the part of the unemployed or the

industrial workers was buttressed by the media. Given the prevalence of anti-strike

messaging, the scope for rallying a majority to forestall retrenchment in the mass

polity was greatly limited.

In the third chapter, I assess the evidence for parallel publics in the late 1930s

and the limits of public tolerance for industrial militance. Based on Gallup surveys,

I find that the mass public, across regional and class lines, had limited tolerance

for industrial militance well before the Reconversion strikes, which are traditionally

considered the proximate cause for the move to Taft-Hartley. This tandem subgroup

variation in opinion is suggestive of an anti-labor bulwark in pre-War mass opinion.

This implies that at the height of its power, and in light of the president’s advocacy

in the court-packing episode, organized labor did not command substantial popular

support.

In the fourth chapter, I directly analyze the states and ask why did states adopt

anti-strike laws and why did states retrench against protective laws they had? I find

support for the claim that the laws were adopted swiftly to contain strikes. Further-

more, I find that the 1938 elections in the North as well as the South ensured the

decline of the right to strike. Partisan changeover is an important variable but is

the intermediary to backlash against the Congress of Industrial Organizations. The

backlash toward the Congress of Industrial Organizations chiefly took the form of

American Federation of Labor resistance and agriculture opposition. The seats con-

trolled by agriculture in the state chambers turned against the Congress of Industrial

Organizations in the elections of 1938 and sought laws to contain the Congress of

Industrial Organizations. In this chapter and throughout the dissertation, I chiefly

attend to the cases of New York, Wisconsin, and Utah because they vary on the de-

pendent variable from full retrenchment to no retrenchment until Taft-Hartley. The

method of di↵erence indicates that disruption-backlash is the best single explanation

for the emergent pattern of United States labor relations. Furthermore, the experience

of these states exemplified the di�culties the union militants faced in the electorate

and mass polity.

The fifth and final chapter observes that the American people more or less

got what they wanted from New Deal labor law development and asks whether the

abridgments of worker liberty implied by broadly de-legitimized rights to strike are

normatively sustainable (i.e. were the proponents of Wagner reform correct in their

insistence that collective bargaining rights were essential to freedom?) and draws

upon work in history and political theory to advance the proposition that even a

minimal conception of citizenship freedom requires a usable right to strike. More-

over, I suggest that the latter-day trend to oligarchic politics is best understood with

reference to shop floor quiescence arising from the decades-old strike restrictions.

Contents

1 Introduction 1

2 What Did the Public Read About Strikes? 33

3 What Did the Public Think of Strikes? 70

4 State-level Origins of Anti-Strike Laws 118

5 Conclusion 174

i

List of Tables

1.1 Restrictive Laws Adopted by the States . . . . . . . . . . . . . . . . . 2

1.2 Summary of the Cases . . . . . . . . . . . . . . . . . . . . . . . . . . 26

1.3 Timeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

2.1 Flint Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

2.2 Hillman and Lewis Coverage . . . . . . . . . . . . . . . . . . . . . . . 50

2.3 Captive Mines Coverage . . . . . . . . . . . . . . . . . . . . . . . . . 51

2.4 Reconversion Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . 51

2.5 Pre-Wagner Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

4.1 Strike and CIO Membership Trends Sources: Bureau of Labor Statis-

tics and Troy and Sheflin (1965) . . . . . . . . . . . . . . . . . . . . . 167

4.2 Penalized Maximum Likelihood Odds-Ratios of Restrictive Law Adop-

tion, 1936-1947, All 48 States . . . . . . . . . . . . . . . . . . . . . . 168

4.3 Penalized Maximum Likelihood Odds-Ratios of Restrictive Law Adop-

tion, 1936-1947, Non-South . . . . . . . . . . . . . . . . . . . . . . . 168

4.4 Marginal E↵ects on Restrictive Law Adoption, United States, 1936-1947169

4.5 Penalized Maximum Likelihood Chow Test, Non-South DV=Restrictive

Law Adoption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

4.6 Odds Ratios, Pre- and Post-1939 . . . . . . . . . . . . . . . . . . . . 170

ii

4.7 OLS Estimates of E↵ects of Republican Seat Gain in 1938 State Elections170

4.8 Sample Restrictive Laws . . . . . . . . . . . . . . . . . . . . . . . . . 171

4.9 Variables and Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

4.10 Variable Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

4.11 Summary of the Cases . . . . . . . . . . . . . . . . . . . . . . . . . . 173

iii

List of Figures

2.1 Strike Coverage Per Annum . . . . . . . . . . . . . . . . . . . . . . . 46

2.2 Strike Coverage and Strike Volume . . . . . . . . . . . . . . . . . . . 47

2.3 Lewis, Hillman, and Murray Coverage . . . . . . . . . . . . . . . . . . 48

2.4 Lewis, Green, and Bridges Coverage . . . . . . . . . . . . . . . . . . . 49

2.5 Pleas and Demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.1 Union Approval, Northeast, Mid-Atlantic, Pacific . . . . . . . . . . . 84

3.2 Union Approval, East Central, West Central . . . . . . . . . . . . . . 85

3.3 Union Approval, South and Southwest, Mountain West . . . . . . . . 86

3.4 Union Approval, Upper and Middle Class . . . . . . . . . . . . . . . . 87

3.5 Union Approval, Lower Class and Unemployed . . . . . . . . . . . . . 88

3.6 John L. Lewis Approval, Northeast, Mid-Atlantic, Pacific . . . . . . . 90

3.7 John L. Lewis Approval, East Central, West Central . . . . . . . . . . 91

3.8 John L. Lewis Approval, South and Southwest, Mountain West . . . . 92

3.9 John L. Lewis Approval, Upper and Middle Class . . . . . . . . . . . 93

3.10 John L. Lewis Approval, Lower Class and Unemployed . . . . . . . . 94

3.11 Strike Ban Approval, Northeast, Mid-Atlantic, Pacific . . . . . . . . . 95

3.12 Strike Ban Approval, East Central, West Central . . . . . . . . . . . 96

3.13 Strike Ban Approval, South and Southwest, Mountain West . . . . . 97

iv

3.14 Strike Ban Approval, Upper and Middle Class . . . . . . . . . . . . . 98

3.15 Strike Ban Approval, Lower Class and Unemployed . . . . . . . . . . 99

3.16 Wagner Act Approval, Northeast, Mid-Atlantic, Pacific . . . . . . . . 100

3.17 Wagner Act Approval, East Central, West Central . . . . . . . . . . . 101

3.18 Wagner Act Approval, South and Southwest, Mountain West . . . . . 102

3.19 Wagner Act Approval, Upper and Middle Class . . . . . . . . . . . . 103

3.20 Wagner Act Approval, Lower Class and Unemployed . . . . . . . . . 104

3.21 Wagner Act Approval, by FDR Approval . . . . . . . . . . . . . . . . 105

3.22 Court-packing Approval, Northeast, Mid-Atlantic, Pacific . . . . . . . 106

3.23 Court-packing Approval, East Central, West Central . . . . . . . . . 107

3.24 Court-packing Approval, South and Southwest, Mountain West . . . 108

3.25 Court-packing Approval, Upper and Middle Class . . . . . . . . . . . 109

3.26 Court-packing Approval, Lower Class and Unemployed . . . . . . . . 110

3.27 Strike Coverage and Union Approval . . . . . . . . . . . . . . . . . . 111

v

Acknowledgements

First, I would like to express my appreciation and admiration for Robert Shapiro,

my advisor and chair of my committee. If I have got half of what is necessary to be a

professional student of politics, it is largely due to Professor Shapiro’s guidance and

patient encouragement. I have also benefitted greatly from the valuable comments

and suggestions of Professor Ira Katznelson and Professor Kimberley Johnson. In

addition, this work has been improved by the suggestions of Professor Justin Phillips

and Professor Todd Gitlin.

I appreciate the assistance of archivists at the Wisconsin State Historical Soci-

ety as well as the Tamiment Institute. Thanks are also due to sta↵ in the Lehman

Suite at Columbia University. I thank the participants at the Columbia American

Politics seminar; the State Policy panel at NPSA; and those at the State Politics in

Historical Perspective seminar at APSA for helpful suggestions and queries.

vi

Dedication

To the memory of:

Althea Kelly

Woodrow Kelly

Patricia Daniel

Theodis Daniel

vii

Chapter 1

Introduction

Motivation

”I think our freedom depends upon maintaining the free right to strike,” said

Senator Robert Taft (R-OH) in concurrence with President Truman’s veto message

regarding the Labor-Management Relations Act of 1947 (Legislative History of the

Labor Management Relations Act, 1653). This claim is a central motivation for my

dissertation. And, in light of this, I pose three central questions: Why did the United

States adopt laws to hamper worker protest despite the fact that worker protest os-

tensibly serves the interest of the majority of citizens? Further, were citizens broadly

in favor or opposed to these restrictions? Finally, what would explain these quick

policy shifts during the 1930s that stand as extant law? Throughout I will refer to

cases of states that adopted Wagner-style (protective) laws and then contemplated

retrenching (restrictive) amendments. Attention to these cases brings the mecha-

nisms of disruption-backlash into sharp relief. These will show that alternate polit-

ical strategies for labor would have failed to forestall Taft-Hartley (i.e. could labor

have coalesced with the farmers; or, should labor have been non-compliant with Taft-

1

Table 1.1: Restrictive Laws Adopted by the States

Law Adopted No Law AdoptedProtective to Restrictive PA (MA) (NY) WI UTRestrictive AL AR AZ CA CO, CT IL IN

DE FL GA IA ID KY MT NCLA KS NE MD ME MI NH NJ NMMN MO MS ND OH NV OK RI SCOR SD TN TX VA VT WA WV WY

Hartley as capital was with Wagner before it?); further, I will argue that Senator

Taft’s, and, before him, Senator Wagner’s, use of the word ’freedom’ is accidentally

illuminating. In the present chapter, I will explain the motivation for revisiting the

New Deal and the topical emphasis on strikes and then I will outline the plan for the

remainder of the dissertation.

Perhaps chief among the motivations for this dissertation is the imbalance of

scholarly attention paid to the states during the New Deal, and particularly from

1937-1947. (More on the periodization in a moment.) The literature, mirroring the

CIO itself, has overwhelmingly focused on the federal level. Theories animated by or

developed in the APD and labor history literature (Greenstone 1977 [1969]; Hacker

and Pierson 2002; Plotke 1996; Katznelson et al. 1993; Farhang and Katznelson 2005;

Schickler and Caughey 2011; Skocpol and Finegold 1982; Skocpol et al. 1990), critical

legal studies literature (Forbath 1991; Gross 1974; Gross 1981; Stone 1981; Tomlins

1985), the class power and class-disruption literature (Domho↵ 1990; Domho↵ and

Webber 2011; Goldfield 1987; Goldfield 1989; Piven and Cloward 1977; Rubin, Gri�n,

and Wallace 1983; Swenson 2004), and the organized labor history literature (Davis

1986; Lichtenstein 2003[1982]; Zieger 1994; Zieger 1995) do not understand the shift

from Wagner to Taft-Hartley as a unitary process - that is, not as a backlash to the

growing power of the industrial unions. Though there is a growing APD literature

2

that takes the states seriously as an object of analysis during the formative 1930s and

1940s (Orren and Skowronek 2004; Weir 2005; Chen 2007; Baylor 2013), the field of

labor relations has been overlooked.

This imbalance in scholarly attention has also privileged top-down as opposed

to bottom-up causal theorizing (cf. Shapiro 2015). The literature does not refute the

pleas of observers that had labor acted di↵erently in other respects, then it would have

gotten a less restrictive labor law regime; contemporary observers and subsequent

students of American political development took labor to task for failing to oppose

segregation, for auto-retrenching against militance, for abandoning the third-party

option too quickly (Milton 1982; Lichtenstein 1998; Moody 1988) among other errors.

This investigation of media, public opinion and state-level lawmaking suggests that

the limits of liberalism might be found in the mass public as opposed to the preferences

of state o�cials or corporate liberals. A top-down theory of the New Deal would

suggest that the Wagner Act might have led to corporatism in the United States as the

Labor Boards promised a ’politicized bargaining’ that would have engendered state-

backed corporatism had Taft-Hartley not been enacted (cf. Lichtenstein 2002). A

bottom-up theory asks whether the public had a great deal of tolerance for a muscular

labor movement; attending to the states asks why were the Taft-Hartley reforms

adopted in the states so quickly after Wagner. In addressing these two questions,

I will point to an understanding of the New Deal as a period where the scope for

a pro-labor revision of the labor relations order was not much greater than it is at

present. Here I am assuming that public opinion plays two roles in a democracy: at

once an object of contestation and as an actor pressing demands.

This investigation of political behavior at the turning point begins by showing

that citizens’ shared an information environment in common that was characterized

3

by framing strikes in opposition to lawful rule. This, in turn, contributed to the

widely-shared public opposition to union militance. The public antipathy toward

strikes generated an electoral backlash at the state level that preceded the Taft-

Hartley backlash. This argument proceeds via a series of tests that are suggestive,

though not decisive. That is, the chapters do not stand alone; but, taken together, the

chapters strongly support the claim that a usable right to strike had been truncated

as a direct result of its having been exercised.

My dissertation exploits the interlude between the adoption of Wagner to the

adoption of the Taft-Hartley amendments in order to explore the origins of the United

States labor relations order. Using this temporal gap, my approach leverages multiple

comparisons to explain the origins of today’s labor laws, rather than relying on the

narrative of one peculiar nation. This is to say that bringing the states into the

discussion increases the N of New Deal labor law outcomes. The structure of this

study enables a broader test regarding the promise of worker and union protest at its

height. And because industrial unions pursued a variety of strategies in di↵erent states

arriving at the same result and because workers faced retrenchment in largely white

states, race can be held constant as well, not having been concretized in economic

arrangements outside the South. In other words, heterogeneous states arrived at the

same result in the absence of labor market segmentation by race.

The core puzzle confronting this dissertation is how to explain why state gov-

ernments quickly reversed the strike rights that were fostered by the Wagner Act.

Given that the states that adopted anti-strike laws were heterogeneous, this disserta-

tion begins by asking what kinds of information citizens had about strikes and then

turns to the opinions expressed by citizens before turning to the states themselves.

People consumed news reports that conditioned their views on strikes which in turn

4

conditioned the policy activity of public o�cials. In order to assess the tenor of news

coverage, I begin with an analysis of the content and volume of strike coverage in

big-city daily newspapers. In order to assess patterns of public opinion, I turn to

an analysis of responses to national surveys.1 And, then, finally, I turn to the states

themselves and draw the link between strikes and retrenching laws.

If the content of newspaper coverage could be construed as favorable, or if there

existed meaningful regional or class conflict in mass opinion, then this argument would

not hold. Further, if the strikes in a given state were not clearly linked to the adoption

of anti-strike laws, the argument in this dissertation would not hold. The argument

in this dissertation stands in tension with received explanations for the sources and

timing of the New Deal reaching the limits of liberalism. If the mass public largely

agreed with restricting the right to strike absent manipulation, then a top-down

account would be incomplete; if states without e↵ectual ethnic cleavages behaved the

same as those with e↵ectual cleavages, then race-based explanations would be likewise

incomplete. The New Deal was not limited by the US entry into the Second World

War; rather, strike rights had been vitiated before then.

This dissertation, then, sets out with the thesis that backlash to the labor

militance explained the adoption of restrictive strike laws at the state level. These

laws would then form the core of the union-hampering provisions that characterize

the extant US laws. This thesis will be supported by the statistical linkage between

strike volume and restrictive law adoption as well as CIO density and Republican

seat gains in the state legislatures. Despite the fact that the CIO was launching

1The tractability of these data are due to the survey response weights devised by Berinsky (2006)and employed in Berinsky and Schickler (2011). The weighted public opinion is reflective of whatmass opinion would have been, given scientifically valid sampling techniques. Nevertheless, becausethe surveys were not conducted in accordance with sound sampling procedures, our analysis cannothinge upon a specific result.

5

strikes that assailed the capitalist structural domination of politics, citizens were not

supportive of that militance. The experience of the states that revised their little

Wagner Acts suggests that the militance alienated essential coalition partners in the

state legislatures. These states adopted provisions that served to placate the coalition

partners disrupted by the CIO. Further, this dissertation sets out with the argument

that the union movement could not have plausibly achieved a better result than

the one it currently enjoys, despite its comparatively weak position. If we include

opportunistic state-level comparisons, we will see that the multiple paths taken at

the state level suggest that alternate courses of action by the organized labor would

not have achieved their aim. This results in two contributions that would elude the

present scholarship on the New Deal: first, attention to the Northern states exposes

a conservative reaction to strikes that created a genealogy for Taft-Hartley; second,

there was no plausible strategy for labor to forestall the onset of restrictive laws.

The Laws

The Wagner Act was the first national law that enshrined workers’ right to

bargain collectively, but was not preemptive legislation so there was space left open

for the states to enact their own policies. The states were able to erect policies

of their own as long as they did not contradict the Wagner Act. The variations

among the states enable the researcher to observe an historical laboratory in the

politics of industrial society. The passage of the Taft-Hartley amendments, though

not preemptive in the technical sense, brought a decisive national answer to the labor

question by divorcing militance from the unions, and by extension, divorcing labor

protest from social or economic protest generally. The swift change in policy did not

apparently run contrary to the wishes of public opinion but the secondary boycott,

hot cargo, and sympathy strike laws raise a tension between freedom and liberty that

6

is normatively ineluctable. But the Wagner Act was not only revolutionary in terms

of enshrining the right to bargain collectively, it also was a most complete intervention

into labor law, which had largely been the province of judicial interpretation guided

by the English common law tradition (cf. Cohen 1948).

Here, I distinguish between labor relations acts and narrower acts as well as pro-

tective and restrictive acts. The labor relations acts are those that established rules

for union recognition and protected the unions from employer interference in a broad

fashion - these are the protective laws; the narrower acts were those that enumer-

ated specific unfair labor practices, typically on the part of employees (particularly

laws against sit-down, secondary and public employee strikes, strikes in violation of

contract, and use of force during strikes) and jurisdictional battles between unions as

well as their internal a↵airs. Here I put the anti-strike laws into one category despite

their distinctive features. It should be noted, though, that these anti-strike provi-

sions were often adopted alongside other measures that anticipated union-hampering

Taft-Hartley provisions (such as provisions prohibiting union political contributions,

right-to-work laws, and limitation on strike funds).

A note on the chronology and geography of the laws. The initial wave of laws in

1937 was pro-union and largely inspired by the Wagner Act. As a result of the statute

and favorable case law, union strength increased and the CIO growth disrupted extant

relations in farm industries as well as old craft unions. Moreover, the sit-down strikes

that brought US Steel and General Motors to the bargaining table seemed to a�rm

the belief that the new unions were a potent force. Thus in 1939 demand for laws on

the part of those who lost as a result of union militance to curb union power increased

and the restrictive wave began. And the focus of the laws shifted from equalizing

relations between employees and owners to an emphasis on restricting union activity

7

in support of the rights of the public.2 The early War years saw no restrictive laws

adopted but, in 1943, a new wave began, particularly in the South, characterized

by laws designed to restrict the internal a↵airs of unions and new organizing. The

attempts on the part of the CIO to use the war as an impetus for new industrial

organizing in the South were the evident cause for this activity. It is worth noting,

however, that this logic would not explain the persistence of Utah’s protective law,

race notwithstanding. That is, war-related industrial growth a↵ected the South and

Southwest states alike but the Utah laws remained relatively more favorable to the

unions.

The New Deal period was crucial in the development of the administrative state

(Skocpol and Finegold 1982; Schiller 2016). And the growth of the labor relations

bureaucracy is one of the defining features of the New Deal program. The statutory

changes in Norris-LaGuardia (the anti-injunction Act) and Wagner (the National

Labor Relations Act) spurred new organizing, particularly in auto, steel, rubber,

and garment unions (Bernstein 1964; Zieger 1994). When these laws were adopted

the industrial unions - the CIO unions - grew. And when the CIO grew it caused

the fracture of the political coalitions that adopted the laws in the first place. But,

importantly, the unions relied on friendly party control of government for the laws

to function: the complaining union in a Board matter controlled neither the decision

to prosecute nor the actual prosecution of the case. Because the state was assigned

the duty of preventing unfair labor practices, the board’s own attorney presented

the evidence in support of the complaint. Also, it is worth noting that the National

Board and the state Boards routinely sought to redress grievances without hearings

2It should be noted that no labor relations act passed during this period defined the term ’public.’Distinguishing citizen qua citizen from citizen qua worker bedevils politicians and students of politicsalike.

8

or sanctions (cf. Herzog 1942).

The laws varied across the regions but the di↵erent regional circumstances im-

plied di↵erent forms of government relief. That is, the more agricultural and rural

south and west were situated in opposition to the industrial north. And this section-

alism did not escape the notice of legislators at the state level (Patterson 1969; Fink

1974).

The labor relations statutes contained many of the provisions of the later Taft-

Hartley, including bans on sit-down strikes, strikes by force, limitations on peaceful

picketing, jurisdictional strikes, sympathy strikes, strikes in violation of agreements,

refusal to transport ’hot’ cargo, coercion, and closed shops. Sit-down strikes were

defined by the peaceful occupation and seizure of industrial work sites by the CIO

unions in steel, auto, and rubber (Bernstein 1969; Fine 1969; Zieger 1997). These were

particularly disruptive because they made the use of replacement workers impossible

without inciting violence. Sympathy strikes are straightforward and ’hot cargo’ strikes

are a particular kind of sympathy strike used by transportation unions (Teamsters)

who refused to deliver goods produced by firms in labor disputes (Interestingly, the

subsequent Teamster leadership would consider sympathy strikes too costly; Ahlquist

and Levi 2013, 55-68 provide a useful summary). Prohibiting jurisdictional strikes

was designed to curb CIO inroads into AFL unions (Millis 1950, 471-4) because the

prohibition served to maintain existing AFL unions in plants and sites where the CIO

might have made further gains. The right-to-work laws compel the unions to deliver

collective benefits to free-riders thus undoing the logic of collective organizing. The

general labor relations acts of the restrictive variety usually combine some or all of

these features into a larger bill, not infrequently with open shop provisions, which

made it illegal for employment to be conditional on union membership.

9

Importantly, the sit-down strike constituted a direct o↵ensive against the owner.

Not only is labor being withheld but also the site of labor is being occupied. Moreover,

the workers in a defensive position on the shop-floor were much more di�cult to

disperse, and attempts at forcible dispersion would entail further capital goods losses

on the part of the owners. The sit-down strike became a tactic used to decisive e↵ect

by CIO unions in steel, auto, and rubber recognition strikes (Fine 1969). Statutory

prohibitions on strikes by force and peaceful picketing follow a similar logic. The

sympathy strike and ’hot’ cargo prohibitions keep unions that represent ’una↵ected’

bargaining units from using outside disputes as a reason for militancy. That is, while

capital was a firm-level actor, labor was a site-level actor. And, indeed, labor leaders

might well have traded Wagner protections for the right of secondary boycott were

they given the choice (cf. Compa 2015, note 5). These kinds of strikes were powerful

because they enabled workers in other areas of the supply chain or geographically

adjacent unions to join in solidaristic resistance to employer directives and capitalist

rule. Though Senator Ives (R-NY) described the secondary boycott as a ”concerted

attempt on the part of a strong union to compel employers to deal with them, even

though the employees of that employer desire to be represented by other unions, or

not to be represented at all” (Legislative History, 1056) In the United States, labor

unions had to strike to organize prior to Norris-LaGuardia and Wagner, so that it

might strike and organize under the law; but prosecuting these strikes for purposes

other than union recognition was politically costly unless, perhaps, the strike might

have been married to broader group claims.

The metamorphosis of labor law from being protective and integral to being

restrictive and contractual3 is a process that occurs in multiple political units simul-

3By ’integral,’ it is meant that the law must shield bargaining rights, because those rights areintegrated with individual rights; by ’contractual,’ it is meant that the wage bargain, however struck,

10

taneously. There was less emphasis on the importance of establishing equal bargaining

power as a first condition of industrial peace, and more on the need to protect the

community against possible excesses by both sides to an existing controversy.

A substantive interest in latter-day economic and political inequality turns my

attention to the New Deal when the extant laws governing private workplace gover-

nance were adopted. The chief goal of the resultant legal regime as codified in the

1947 Labor Management Relations Act (informally, the Taft-Hartley Act) was the

containment of disruptive militance. And since the adoption of Taft-Hartley, union

membership and coverage has diminished and, importantly, strikes have diminished

while income and wealth inequality has grown in the United States (Rosenfeld 2014,

Chapter 4). The decline in strikes is the result of the distinctive laws of the United

States with respect to hiring replacement workers (Ibid 87-9). This work, though re-

liant on Jacob Rosenfeld’s insights, is distinctive in that I understand that the decline

of strikes was not purely the result of the increased use of replacement workers - or

organizing tactics, generally; all the successful employer resistance tactics were e↵ec-

tive because the unions cannot engage in sympathy or secondary strikes. Ultimately,

though, the retrenchment against worker protest was a democratic result and, indeed,

was a fairly popular result across the nation.

Why did some states adopt laws that restricted unions’ right to strike while

others did not? Why were citizens unwilling to ratify radical worker protest? These

related questions expose the distinctive character of United States labor relations

order. In the United States, the workplace is the domain of the employer and workers

are meant to obey the dictates of owners and managers. Workers can be fired for

any reason, and, therefore are in an insecure position. Unions tend to increase the

is an agreement between free equals.

11

security of workers (Freeman and Medo↵ 1983, Ch. 2). However, strike rights, not

union membership, is the focus of this dissertation because the right to strike is more

important than the right to organize. To wit, France has an even lower union density

than the United States; yet more robust statutory and regulatory protections exist

for workers in France than the United States because of the legal capacity for worker

protest and strikes in public, and more importantly, at the site of production.

Definitions and Guiding Assumptions

Following Karen Orren (1995, 387), I define labor relations as the pattern(s) of

human interaction – characterized by compensation, subjugation, and discharge – by

which a society organizes purposeful activity of any notable scope. The labor relations

laws are those that govern the relations of unions, workers, and employers in the

workplace and the spaces adjacent to it. The extant laws are largely intact from the

1947 Taft-Hartley amendments to the 1937 Wagner Act – the first national level labor

relations law. The Wagner Act, the informal name for the National Labor Relations

Act, was the first of its kind and was under a cloud of constitutional suspicion from its

adoption. (Indeed, the Supreme Court ruling the collective bargaining provisions of

the National Industrial Recovery Act as unconstitutional precipitated its adoption.)

These laws were themselves adopted because of disruptive militance on the part of

the CIO. Throughout the empirical chapters, we will explore three industrial states

to lend historical basis to the pro↵ered mechanism of union-backlash.

The adoption of the restrictive amendments to the National Labor Relations

Act was connected to the subsequent decline of the American (private-sector) unions.

And it is possible that the states were decisive contributors to the canalization of

the American labor movement; indeed, the manifold provisions of the Taft-Hartley

amendments are drawn from state laws. The Taft-Hartley amendment to Section

12

8(b)(4)(B) banning ’hot cargo’ strikes was borrowed from several state legislatures in-

cluding Delaware, Florida, California and Minnesota (12 total). State-level provisions,

including Louisiana, Pennsylvania and Iowa (10 total), antedated the amendment to

Section 8(d)(4) that bans strikes in violation of contract. And, of course, Section

14(b), appended to the Wagner Act, which allows states to enact right-to-work laws

(7 total), was enacted after several states had already enacted them - all Southern

states (Alabama, Arkansas, Florida and Texas) save Arizona, Nebraska, and South

Dakota. To wit, Taft intervened on the Senate floor: ”I should like to call attention

to the fact that the [prohibition on picketing] amendment is practically contained in

the Wisconsin Labor Relations Act and has been used in Wisconsin as a means not

only of preventing the coercion of employees but also of attempting bringing such

action as can be brought by administrative law, to end mass picketing [...] it seems

the simplest matter to use the same words in this law...” (Legislative History, 1032).

I use the states to evaluate counterfactuals because inference in historical social

science faces an obstacle in that

History is not rule governed and it knows no su�cient causes [...] Thehistorians of the future, who will know how things turned out, will have apowerful aid to understanding, not why they had to turn out in that way,but why in fact they did: that is, they will observe in the laboratory ofevents the evidence of determination, not in its sense as a rule-governedlaw but in its sense of the setting of limits and the exerting of pressures.For historical explanation discloses not how history must have eventuatedbut why it eventuated in this way and not in other ways; that process isnot arbitrary but has its own regularity and rationality...characterized bydeterminate relations and by a particular logic of process (EP Thompson2001 [1978], 457-8 emphasis his).

This dissertation takes up Thompson’s challenge in two ways: first, by gen-

eralizing the politics of retrenchment against organized labor at its apogee; second,

by exploring the limits of liberalism aside from those identified by the literature on

13

national elites and examining the contours of public opinion alongside contempora-

neous newspaper coverage. Ultimately, the limits of New Deal liberalism were met

in the legislative chambers, and, therefore, the electoral and attitudinal connection is

relevant. This dissertation will examine states with muted ethnic segmentation and

varying levels of militance as well as a an exemplar context wherein the labor unions

sought to forge a partisan coalition outside the Democratic Party and, therefore, point

toward the conclusion that the unions could not have forestalled the present-day anti-

strike regime.

Taking this counterfactual seriously does not imply the researcher is being se-

duced by ’the false promises of hard times’ (Gourevitch 1986); namely the belief that

the short term weakness of typically dominant actors and the ’hard times’ coalitions

that form will endure after the subsequent recovery. Combined membership of the

CIO and AFL was never an impressive figure when compared to the total employable

in the United States. However, it was important to note that the heavy manufactur-

ing industries and the core industries of the economy - motors, electrical equipment,

steel, coal, rubber, glass, and textiles - were all bowing to union control in the opinion

of business observers (Witte Papers 1937, Box 116, Folder 2).

Though the labor movement has been pacified in numerous ways by the Taft-

Hartley amendments, the withholding of labor is paramount. Because the coercive

apparatus will be applied against unions that strike illegally, the only legitimate

strikes are those that concern wages and shopfloor conditions and must be limited to

that site. Capital operates across many firms while labor is only allowed to operate in

one firm at a time. Thus, the site-level division of labor combined with the industry-

level division of firms resulted in a labor law regime wherein labor could not operate

solidaristically while owners could (Tomlins 1985, Ch. 7; Lambert 2005, 122-4).

14

It would be tempting to situate counterfactuals4 in the rich drama of the Taft-

Hartley veto override in the Senate. Had Wagner (D-NY) made it o↵ his deathbed

and Thomas (D-UT) taken a steamboat back from Europe, the veto would have

only been five votes shy of sustention. Hatch (D-NM) was the only non-Southern

Democrat to vote for Wagner and vote to override Truman’s veto; White (R-ME)

voted for both Wagner and the veto override. The remaining seats were in the hands

of alarmed Dixiecrats or hostile Republicans, so positing a world where the veto would

have been sustained is mistaken. But, even if it were not a mistake to imagine Taft-

Hartley being repudiated by the Senate (the House was out of reach for the pro-labor

side), it would be less inferentially fruitful to use one historical narrative than to use

several concurrent, commensurable narratives.

The history of labor relations before the Great Depression was one of private

warfare marked by employer repression and law enforcement intervention. Before

the 1930s, the federal government tended to act as a strikebreaker if it intervened.

The 1935 Wagner Act constituted a sharp, though temporary, reversal in policy and

institutional behavior. And, importantly, it seemed to indicate an increase in the

power of unions and the working class generally. But this was a chimera: labor unions

and their liberal allies were engaged in a race against time, seeking to construct a

durable political-economic place for organized labor before the capitalist opposition

reorganized and retrenched (Schickler and Caughey 2011, 146). The attempt to win

the race was doomed because labor could not form a majority on its own, and the

only likely ally with the capacity for e�cacy in the political space was the farm

bloc. Although they could form transient legislative majorities on the state level, the

4A key principle of counterfactual analysis is that as little of the world should change as possiblein the alternate case under comparison (Weber 1949; Lewis 1973; Fearon 1991). My approach herestructures the counterfactuals without resort to a conception of an alternate non-existent world, forthe states themselves o↵er comparisons wherein the causal antecedent is absent or di↵erent.

15

farmer-labor capacity to form a national alliance as John L. Lewis had wanted, was

limited by the by antagonism wrought by labor stoppages. These were the selfsame

reasons that the farmer-labor alliance was transient at the state level, as we will see

in Wisconsin.

The transition from protection to restriction turns out to be an unintended

consequence of purposeful behavior. The present-day labor laws are the residue of

New Deal struggle and backlash; the same militance that won the Wagner gains

ensured the Taft-Hartley losses. Social, particularly political, inquiry struggles with

such outcomes, as it is di�cult to explain them beyond noting that an actor or group

made a miscalculation or su↵ered from a misperception. As Jervis (1997, 128) puts

it, ”one change leads to others, but these neither reinforce nor dampen the initial

one. They move the system sideways, so to speak.” The CIO strikes led to a negative

reaction in the mass public and among liberal coalition members, which led to the

reaction of restrictive law adoption, the provisions of which structure the extant

labor relations order, which is itself quite distinct from the prevailing ex ante order.

This dissertation, then, does not call the policy feedback and retrenchment literature

into question directly, because actors in those theories are typically motivated by a

previous result at a later juncture (Skocpol 1979, for example) or are conditioned

within institutional structures and historical paths (as in Pierson 1996, for example).

In the period under study, the backlash was immediate and politicians immune to

institutional forces in labor relations enacted the retrenchment.

Implicit in the backlash concept is the notion that the mass polity held unions

in disfavor. This empirical question will be explored as it has been in previous ef-

forts (Bok and Dunlop 1970; Parenti 1986; Schmidt 1993; Flynn 2000; Martin 2004;

Carreiro 2005; Amenta et al. 2009, for example). Distinct from these e↵orts, this

16

dissertation connects the mass intolerance toward worker protest buttressed by media

coverage to the origins of the labor relations institutional framework. Anticipating

a theme of the civil rights movement and the women’s rights movements, the mass

public had little sympathy for disruptive or unconventional protest tactics. This is

necessarily connected to the media coverage of union militance, because mass opinion

is, subject to exposure assumptions, mediated (Iyengar and Kinder 1987; Zaller 1992;

Kellstedt 2003, for example). Put another way, people did not form opinions about

strikes solely from their direct experience of labor stoppages; they relied more upon

news coverage. The linkages between media, opinion, and state-level policymaking

are not drawn directly here, particularly that between media and opinion. Rather,

this dissertation is guided by the assumption that the observed mass antipathy to

strikes was informed by the prevailing unfavorable coverage of strike actions.

In going from Wagner to Taft-Hartley, the United States foreclosed the pos-

sibility of an e↵ectual labor-left movement in politics. The limited construction of

legitimate union militance under restrictive laws – laws that inhibit union or worker

militance or organizing – is directly connected to the rise of concession bargaining and

the decline of unions in the latter part of the twentieth century (Rubin et al. 1983;

Lambert 2005; Rosenfeld 2014). Without the broad power to strike, the workers can-

not resist being denuded: militance that ensnares the site of production is essential

to compel intransigent ownership to engage in bargaining.5 In principle, the union

without the ability to withhold labor is unable to threaten the owners’ equity stake;

moreover, as a general principle, to the extent that the scope of legitimate militance

is limited, the impact of work stoppage is limited. As Estlund points out

5More subtly, the strike power is essential to the capacity for unions to provide a value-addedservice to members. Thus, restraining strikes rightly diminishes the worker’s rational commitmentto or willingness to join the union (cf. Moore and Newman 1985).

17

Unions cannot thrive if their employers do not. But employers can thrivewithout unions and many would prefer to see unions die out [...] Whileorganized labor has mounted several major e↵orts at labor law reform inthe past thirty years, employers have made almost no such e↵orts since1959. For the most part, employers who oppose unions and collectivebargaining are willing to bide their time in the political process, battingdown periodic reform proposals that might tip the scales in unions’ favorand watching union strength ebb away (2002, 1543-4).

Scope

Some states adopted laws that were akin to the Wagner Act in 1937, shortly

after the Supreme Court decided in favor of its constitutionality. Most of those states

in addition to states that had no labor relations statutes then adopted retrenching laws

akin and prior to Taft-Hartley after the 1938 elections. That is, some states adopted

Taft-Hartley precursors without a ’baby Wagner’ law, whilst some states passed no

labor relations statute. The Taft-Hartley Act draws its anti-strike provisions from

state laws adopted in the preceding decade. As several states adopted protective laws

modeled after Wagner and then quickly retrenched with restrictive statues, a closer

look at the states is indicated.

Table 1.1 depicts the states aligned on the key dependent variable of the disser-

tation: whether a restrictive strike law was adopted in a given state; and whether the

states that adopted protective laws patterned after the Wagner Act adopted amend-

ing laws anticipating Taft-Hartley restrictions. The inferential leveraged gained from

these multiple comparisons allows an exploration of all credible counterfactuals: by

taking seriously the possibility of a third-party farmer-labor coalition (Zieger 1988,

Chapter 4); or by taking seriously the call for non-compliance issued by the left-led

unions (Rosswurm 1992; Stepan-Norris and Zeitlin 1995; Zeitlin and Weyher 2001);

or by taking seriously the spatial unevenness of militance (i.e. might labor have been

18

more determined in their resistance to owners?).6

So much for the spatial scope. The temporal scope is bounded by the Supreme

Court upholding the Wagner Act in 1937 and the adoption of Taft-Hartley in 1947

which signaled the crystallizing of the United States labor relations order, which

has been intact on a federal level since the Eisenhower administration (Orren 1995;

Weatherford 2014). Moreover, since the laws governing the spaces and time where

most citizens spend most of their lives enable tyranny, the tyranny they endure is

necessarily of substantive interest to students of politics (Doellgast 2013, for a recent

example). The unquestioned prerogative of management to control the American

shopfloor (which nowadays is more often a nursery of cubicles than a mass production

enterprise) lays the foundation for the ills confronting American democracy (Winters

and Page 2009; Frymer 2010). Under the aegis of the law, agents for corporate

shareholders oversee a demoralized and insecure workforce; a workforce that translates

into an insecure mass overseen by an ensconced elite (Estlund 2002; Doellgast 2013).

The cowed workforce of the present stands in sharp contrast to the resistance o↵ered

by the workers of the young Committee for Industrial Organizing.

More so than ever before, the years from the New Deal through Reconversion

(converting the war economy to peacetime, chiefly by lifting wage and price controls)

constituted an enduring shift in authority in labor relations (Orren and Skowronek

2004, Chapter 4). Moreover, this period is critical because the Depression and World

6An alternate potential counterfactual concerns the South: many lamented the failure of Opera-tion Dixie and historians have aptly noted that the early civil rights demanders were activated by theCIO’s e↵orts to organize the South in the mid-1940s (Gri�th 1988; Honey 1993, 1-44, for example).But, a crucial lesson from this dissertation is that US organized labor would have failed even if therewere not a racial segmentation in market and society. Wisconsin, among others, was a lily-whitestate that adopted restrictive legislation. Moreover, the confluence of newspaper coverage and theconvergence of subgroup opinion observed later in this dissertation suggests that the prospects forlaws that would allow labor to flourish were unlikely to sustain popularity.

19

War II spurts raised union membership to unprecedented levels - and this pattern

might have led an observer to conclude that unions would become a permanently

influential socio-political actor. However the laws arising from the selfsame period

foreclosed this outcome (Freeman 1997, 15).

The present focus on the strike laws stands in contrast to canonical work on

United States organized labor which has focused on organizing defeats (Freeman and

Medo↵ 1984; Goldfield 1987; Farber and Western 2001). I argue strikes are more

important for understanding the e�cacy of unions in light of the logic of industrial

bargaining and historical experience; that is, without the capacity to collectively

withhold labor, the workers are in a weak bargaining position. More pointedly, the

right to withhold labor is an essential component of e↵ectual collective bargaining.

If workers can be replaced for striking - or must wait years for legal remedy - then

their bargaining position is untenable. Prior to Wagner, the strikers were met with

violence; after Wagner, the strikers were met with replacement (Klare 1978; Rogers

1990; Estlund 2002; Rosenfeld 2014). Though the lot of the wage-earner has improved;

the bargaining power, and, by extension, the political power of the wage-earner has

not.

If bargaining is contentious, then the resolve of the contestants is a crucial de-

terminant of success, and, more subtly, the expectations of resolve are determinative

of strategy. A relevant condition that a↵ects the resolve of the contestants is the polit-

ical outlook of the statehouse. Sympathetic governance directly strengthens resolve.

And partisanship does not explain much in this respect. Democrat Herbert Lehman

and Democrat George Earle, governors of New York and Pennsylvania respectively in

1937 took opposite tacks toward industrial union mobilization during the New Deal.

Governor Earle was willing to call out the National Guard as strikebreakers while

20

Governor Lehman repeatedly refused employer appeals (Davin 2000, 278-80; Kim

2006, 31).7

Though the Taft-Hartley provisions did not amount to being the ”slave labor

law” that then CIO president Philip Murray prophesied (cf. Lichtenstein 1997), it

did su�ce to narrow the scope of militance to when contracts were up and could only

be about wage-and-hour or shop floor conditions. This is in stark contrast to the

multifarious causes for workplace stoppages during the New Deal. During the 1930s,

the industrial unions struck in furtherance of recognition; having won recognition

in the late 1930s the unions struck in furtherance of bargaining gains; during the

1940s, particularly after the war, the preponderance of strikes were related to control

of the shop floor and in furtherance of racial segmentation of labor. Despite the

heterogeneity of motivations, we can be confident that the unionists won whatever

they won because they fought resolutely and confronted repression in a solidaristic

fashion. And, crucially, the e�cacy of the strikes was dependent on the coercive

apparatus not having been readily deployed against them, whether before or during

the Second World War.

The result of the New Deal was the erosion of union bargaining power and,

consequently, political power. Because the threat of the stoppage is the origin of

bargaining leverage for the worker, then the threat must have some credibility. The

laws inherited from this period that narrowly circumscribed the right to strike min-

imized the cost of a workplace stoppage to the owner. The resultant strike weapon

was too predictable and was nearly toothless because the employer could anticipate

strikes e↵ectively and maintain production; consequently, workers were left with little

7Though it should be noted that there was incidence of National Guard insubordination; to wit,guardsmen had alerted unionists and radicals to the troop movements (Vallely 1989, 112-3).

21

in the way of positive means to advance programmatic goals within the workplace

and beyond.

Grounding Examples

Throughout the dissertation, I apply the overarching argument to specific se-

lected industrial states (see Table 1.2) that vary in whether and how much retrench-

ment there was. These states vary along important axes of comparison though the ab-

sence of Jim Crow laws is common among them. These parentheses indicate variation

along the dependent variable that is not quite as far as the full turn to restriction; that

is, New York and Massachusetts adopted restrictive laws that limited jurisdictional

strikes but did not restrict many e↵ective strike tactics. So then, Pennsylvania and

Wisconsin retrenched, New York and Massachusetts partially retrenched and Utah

did not retrench. I do not select on the dependent variable because Utah’s adoption

of a retrenching law was adopted after the preemptive retrenching law was passed

by the Congress. The political mechanics of disruption-backlash are exemplified by

the states and common among the retrenching states, aside from New York, were

electoral rules that dictate a two-party system. That is, the rules of the game are evi-

dently the dominant reason that organized worker resistance was and remains stifled.

There did remain one potential avenue for labor to influence politics without having

to be mediated through the Democratic party: namely, through the construction of

a farmer-labor coalition on a national level.

I focus separately on the cases of New York, Wisconsin, and Utah because

these states form the closest analogues to the federal transition from Wagner to Taft-

Hartley. An assumption of equivalence among the actors within the states relied upon

here is based on a formal conception of class interest and formation that identifies the

shared utility function of individuals in similar straits as the definitional feature of

22

class membership. To the extent that the utility is maximized by the same strategies,

those individuals are in the same class. For the theoretical purposes of this project,

I treat classes as exclusive categories.

Di↵erent sectors were predominant in the cases: the building and meatpacking

trades dominated in Wisconsin and the garment trade was predominant in New York

(WLRB Report 1937; NYS Annual Report 1937). In Utah, the mining industry was

ascendant (McCormick and Silito 2011; Powell 1977). There were important institu-

tional variations: particularly in terms of the lack of party discipline in the Wisconsin

legislature and in terms of the deep connection between FDR’s administration and

the American Labor Party in New York. And, importantly, the Wagner Act preceded

the adoption the ’little Wagner Acts.’

The labor relations activities of New York, Wisconsin and Utah help frame an

analysis of labor law adoption in the states during the New Deal. An important goal

is to point the way toward a conceptual understanding of disruption-backlash. This

study looks to these states because they are heterogeneous along the axes of compar-

ison introduced in the above and because they are directly analogous to the nation.

Further, these cases are selected because the union security laws created the most

active Labor Boards in the nation; thus, these were crucial battlegrounds during the

late 1930s. More subtly, both New York and Wisconsin had professional legislatures.

Implicit in the logic of comparison is an assumption of equivalence among the cases.

The actors in the cases need to be commensurable and the realm of contestation needs

to be comparable. These states are useful for framing the discussion because they

traveled the road from Wagner to Taft-Hartley before the rest of the nation.

The labor relations acts provided a machinery for so-called employer unfair labor

practices through administrative review. The Massachusetts, New York, Pennsylva-

23

nia, Wisconsin, and Utah laws of 1937 and the Rhode Island law of 1941 established

’little Wagner’ Acts. In the cases of the early protective laws, the states formed La-

bor Boards to make the legislative policy e↵ective. The New York Board was, by far,

the most active, handling as many cases per annum as the other Boards combined

(Killingsworth 1948, 26).

The New York Labor Relations Act passed in April 1937 was the product of

a Democratic governorship lead by FDR’s lieutenant, Herbert Lehman. The Doyle-

Neustein bill passed the Assembly 101-46 and the Senate 46-2. As Governor Lehman

recalls, 1936 was the only year that brought a unified government in New York during

his tenure (Oral History Project). The wave of Depression strikes that swept the

nation were similarly widespread in New York. But these strikes were largely in the

garment industry.

The Wisconsin Labor Relations act was adopted in May during the 1937 legisla-

tive session shortly after the ’switch in time that saved nine.’ The bill, though, was

first introduced during the 1935 legislative session after years of violent strikes. The

bill failed in the upper chamber of the state legislature and this became an issue in

the 1936 elections (Blue Book 1937, 17-70). The bill passed the Assembly 66-30 but

only passed in the Senate by two votes 29-27. Once adopted, it established a state

labor relations board which actively promoted collective bargaining through imposing

fines on employers found to engage in unfair labor practices. And, like the Wagner

Act, the unfair labor practices of employers were the only ones defined by statute and

regulation.

Utah adopted its baby Wagner law in April 1937, having passed the Assembly

43-2 and the Senate 21-8, and was signed into law by the Democratic Governor Henry

Blood. It did not adopt a restrictive statute until 1947 after the federal government

24

had adopted the national law. Interestingly, it was Herbert Maw, a liberal Democrat,

who signed the retrenchment bill. Perhaps it was an attempt to stave o↵ defeat in

his rematch with the Republican, Bracken Lee (more on this later).

In a formal historical narrative analysis, causal process tracing would be used to

compare the adoption of the State Labor Relations Acts (SLRAs) and the conservative

amendments in subsequent legislative sessions. For the purposes of this dissertation, it

will be su�cient to note the stated positions of unions and politicians while accounting

for the immediate upshots of the SLRAs. That is, an examination of the passage of

the bills and their later amendment suggests the unifying theme of CIO disruption

as being determinative as against economic factors or alternate political theories;

however, a more detailed tracing of historical events would be necessary to examine

the political behavior results of the AFL/CIO rivalry.

The method of di↵erence, though, is su�cient to rule out competing theories:

herein I compare cases that di↵er on variables on concern but are similar on others.

In this approach, the key is to find an explanatory variable along which the cases

di↵er that is itself congruent with the dependent variable (cf. Brady and Collier

2010, 337-8). There were strikes before and after the constitutionality of Wagner or

the adoption of SLRAs, which militates against class-disruption theory. Economic

variables were heterogeneous across the cases; though the Depression ravaged every

political unit, Wisconsin, New York, and the nation as a whole had variable industry

composition and factors of production. In Wisconsin, for example, land was a more

important factor than in New York or the nation generally. Yet, in the face of this

heterogeneity, the same kinds of statutes were adopted and then amended in similar

ways; so this militates against a chiefly economic basis for policy change.

In Table 1.2 above, it is evident that the count of owned farm homes distin-

25

Table 1.2: Summary of the Cases

Wisc. New York Penn. Utah Mass.Protective Law 1937 1937 1937 1937 1937Restrictive Law 1939 (1940) 1939 1947 (1941)Change to GOP Yes No Yes No No

Strike Volume Rank 8 6 4 41 23Farm Owner Rank 14 27 28 13 15

guishes Wisconsin and Pennsylvania. However, consonant with the statistical results

from the fourth chapter, the rural delegations in the statehouses were the main blocs

in retrenching legislation. Because the farmer-labor possibility was a threatening one

to the status quo this dissertation will focus on Wisconsin and treat Pennsylvania as

a site for opportunist comparison.

The farmer-labor divide was historically constructed; the farmers had a his-

tory of policy success in the US during the latter part of the 19th century (Prasad

2012, Chs. 1-2). This meant that labor growth politically constituted a force that

diminished the preeminence of agricultural interests. This served as a sti↵ barrier to

cooperation in itself. The American state was non-interventionist for labor but was

not for farm interests (Boudreaux and DiLorenzo 1993; Prasad 2012). And, more sub-

tly, the American farmer who owned his enterprise was more inclined to view himself

as a capitalist than as a laborer. In any event, agricultural interests were given to

viewing wage gains warily; for example, Milo Swanton, a leader and organizer of farm

owners concluded that: ”It is true in a country like ours, with surplus farm produc-

tion, that, when the farmers are prosperous the rest of society is prosperous. This is

true because a surplus of farm product cannot be sold at good prices to unprosperous

customers. The converse of this statement is not, however, equally true.” (Swanton

1936, ”Mutual Relations of Agriculture, Labor, and Industry” WCAC Files, Box 8,

Folder 1). That is, many might well prosper when the farmer does not, but the agri-

26

culture interest cannot prosper when others do not. In a separate address, ”Butter

is the same in price as twenty-five years ago, while the wages of skilled workmen are,

roughly, fifty percent higher. Can the farmer help to boost wages which he has to pay,

while his own income stagnates, or goes lower?” (Ibid 1937, ”Address to Milwaukee

Chamber of Commerce” WCAC Files, Box 8, Folder 5).

A notion of the unions as invader figured prominently in the descriptions of

John L. Lewis’ attempts to unionize the dairy industry after his break with the CIO

(WSHS, Milo Swanton Papers). Lewis had an ambition to put a national farmer-

labor majority together8 that was frustrated from the outset. For example, ”Control

of the food supply is a weapon. John L. Lewis knows that if he controls the nation’s

milk supply, he controls the destiny of the people” (Milo Swanton Papers, ”Hoard’s

Dairyman” June, 1942). Because the farmer is historically prior to the laborer, agri-

culture concerns were unwilling to join with labor unions as a junior partner. 9 Upon

winning in the 1936 elections, the farmer-labor majority in Wisconsin saw its role as

concentrating ”on pursuing legislation to the left of the New Deal [...] Farm tenantry

must be attacked by a resettlement program. It is important we support civil liberties

as a means of opposing fascism. Finally, we must press for cooperative ownership on

a wider scale than otherwise attempted” (Amlie Papers, ”Joint Meeting of Wisconsin

and Minnesota Farmer-Labor Parties”). The socialistic and statesmanlike vision of

the progressive leadership was at odds with the interests of the owners of dairies.

Plan of the Dissertation

8In the 1930s, more than a third of the population of Wisconsin lived on farms; about a fifth ofUtah’s residents lived on farms; and about a tenth of New York’s population did (Statistical Abstract1937, Table 10).

9Alternately, perhaps agriculture went to war with the Axis alongside the national government.That is, in the same way that patriotic concerns influenced unionists’ compliance with no-strikepledges (cf. Gerstle 1991), farmers may well have been willing to subsume their concerns to the ware↵ort.

27

These multiple comparisons lend insight into the political dynamics of public

choice during the New Deal, holding racial segmentation constant while envisioning

the process of legal change as a bottom-up as well as top-down process. The legal

outcome of strike restriction was a product of mass opinion (as measured by ballots

and surveys) which in turn was embedded in a media context. In short, the prevailing

coverage of strikes facilitated citizens’ unfavorable opinions of strikers and strikes;

this, in turn, led to anti-strike laws being adopted. This dissertation unravels the

connection in the following chapters. Conceptually, these chapters proceed from an

examination of the information people had to an examination of people’s attitudes

and then an examination of policy outcomes in the states. Insight from case studies

of New York, Utah, and Wisconsin will be interspersed throughout.

The next chapter, ”What Did People Read About Strikes,” I ask when Ameri-

cans consumed news about strikes, what did they ingest? How in terms of quantity

and tenor did the news portray strikes? I will provide counts of strike and related

articles as measures of salience; examine the frames of coverage of strikes as disrup-

tion; I shall examine the framing of unions as racketeering organizations as opposed

to welfare organizations; and I shall examine the framing of John L. Lewis, the CIO

labor leader. These frames are chosen in light of the lines of attack chosen by union

adversaries. These frames stand in contrast to those employed by the CIO, which

tended to emphasize shopfloor hazards and wage and hour disputes.

I argue that the media coverage of major strikes in newspapers with wide cir-

culation consistently emphasized the inconvenience or disorder resulting from strikes

without considering the basis for worker demands. Indeed, the use of the word ’de-

mand’ as opposed to ’grievance’ or ’plea,’ suggests that the striking unions are a pos-

itive threat. Widespread negative coverage of strikes was connected to the backlash

28

against industrial union growth, which was impossible without militance. Moreover,

the growing substitution of the word ’management’ for ’capital’ or ’owners’ served to

foster the impression that class interests were not at stake in labor relations. I am not

the first to focus on the content of media coverage of strikes (Parenti 1986; Rondinone

2010, for example) or successful business e↵orts to frame labor relations in a fashion

that obscures class conflict (Fones-Wolf 1994; Phillips-Fein 2011, for example), but

the marriage of studies of the several states and empirical study of media and surveys

during the New Deal in order to incorporate mass opinion into the study of the limits

of liberalism is original.

In the third chapter, ”What Did People Think of Strikes?” I turn to the public

who defined the boundaries of acceptable militance. The elections of 1938 and then

1946 returned majorities to legislative chambers for retrenching laws. The mass pub-

lic, generally speaking, had (and still does have) little tolerance for opinion leadership

from radical militants (Page, Shapiro and Dempsey 1987, 37). Implicit in describing

the speedy retrenchment as a ’backlash,’ is the sense that public opinion indicated

disfavor for unions, or at least union militance. Works in the interest group canon

(Key 1964; Truman 1971) emphasize that the esteem in which the mass public holds

interest groups is directly related to their capacity for political e�cacy. The public

opinion literature that utilizes relevant polling from the period (Verba and Schloz-

man 1977; Berinsky and Schickler 2006; Panagopoulos and Francia 2008; Schickler

and Caughey 2011) has broadly covered this ground before, but the aggregate stabil-

ity of opinion is left unconnected to the development of the labor relations framework.

In contrast to other works which emphasize the 1940s retrenchment, my dissertation

emphasizes the importance of the 1938 elections in the decline of New Deal liberalism;

in contrast to work that emphasizes the importance of the 1938 elections in the elec-

29

tion of conservatives (Patterson 1968; Katznelson and collaborators), my dissertation

finds the limits of liberalism in the North and in overwhelmingly white states.

In the fourth chapter, ”State-level Origins of Anti-strike Laws” describes the

labor laws and proceeds to draw a clear link between union militance and the adoption

of laws designed to neuter resistance. This section will introduce a data set of state-

years from 1937 and 1947, for the reasons noted above. The structured comparison

between the states that adopted protective statutes and amended them and the states

that did not adopt a state law inform the case studies. Furthermore, placing the

origins of the backlash in 1937 casts the 1938 elections in a significant light and

this motivates a turn to studying mass opinion during this period. The viability of

the disruption-backlash explanation is contingent on its place in the 1938 elections,

for the retrenchment against the Wagner Act began at the state-level enacted by

politicians elected in the wake of militance. I find a clear relationship between state-

level militance and Republican seat gains in 1938 statehouses outside the South. The

cases referred to throughout are treated in this chapter in order to exemplify the logic

of political backlash to strikes.

The fifth and concluding chapter takes the findings and non-findings together

and contemplates the implications for union revival alongside managerial domination

of the workplace. The death of the American labor stoppage was an ineluctable, if

delayed, result of the anti-strike laws passed in the 1930s and 40s. Public opinion,

arguably against its own interest, generally favored pro-business strike laws. Interests

are a treacherous topic in terms of imputing them onto individuals on the basis of

supposed commonality. However, observers at least since Freeman and Medo↵ (1984,

251), have seen that the decline of unions and workers as counterpoised groups versus

business would carry implications for wealth inequality and political oligarchy.

30

A distinctive benefit of the mixture of methods deployed in this dissertation is

the resultant insight into whether a di↵erent historical path might have been taken

and why those alternate paths were occluded. Building o↵ the literature that iden-

tifies US laws as the best explanation for its distinctively commoditized labor force

and small welfare state (Esping-Andersen 1990; Pierson 1996; Western 1997; Hall and

Soskice 2001; Hall and Thelen 2009, for example), and the decline of labor stoppages

as the chief source of labor decline (cf. Rosenfeld 2014), I investigate the causes of

restrictive law adoption at their historical origin revealing the proximate cause to be

a reaction to disruptive strikes. The logic of this claim, investigated in the cases,

reveals that organized labor needed coalition partners in order to secure a majority

but this was unrealistic. Labor could not form a majority coalition centered on its

interest in broad socio-political reform because there were bulwarks in mass opinion

against worker radicalism and militance that were buttressed by news media. Over

the course of US industrial and post-industrial society, the political power of unions

has been minimal compared to their own programmatic aspirations or their counter-

parts in other a✏uent polities. Their decline has been decades long but it can be

traced back to labor laws adopted during the turbulent 1930s.

31

Table 1.3: Timeline

January 1935 Textile organizing strikes across the nationMay 1935 Norris-LaGuardia collective bargaining ruled

unconstitutionalJune 1935 Wagner Act adoptedOctober 1935 Congress for Industrial Organization leaves AFLNovember 1935 - March 1936 Rubber Workers Organizing Committee Akron

sit-down strikesApril 1936 Goodyear accepts RWOC demands

save recognitionDecember 1936-February 1937 United Auto Workers-CIO Flint sit-down strikeFebruary 1937 General Motors recognizes the UAWFebruary 1937 Supreme Court Enlargement Bill introduced

in SenateMarch 1937 US Steel recognizes the CIOApril 1937 Supreme Court upholds Wagner ActApril 1937 Utah adopts ’little Wagner Act’April 1937 Wisconsin ’little Wagner Act’ adoptedMay 1937 New York ’little Wagner Act’ adoptedJune 1937 Pennsylvania ’little Wagner Act’ adoptedAugust 1937 Massachusetts ’little Wagner Act’ adoptedNovember 1937 CIO becomes Congress of Industrial OrganizationsApril 1937 CIO organizes dairy cooperatives in Wisconsin

and across MidwestJanuary 1938 Onset of ’Roosevelt recession’November 1938 Republicans gain in mid-term electionMay 1939 Wisconsin ’Labor Peace’ Act adoptedJune 1939 Pennsylvania ’Labor Peace’ Act adoptedMarch 1940 Smith Committee hearingsNovember 1940 John L. Lewis resigns CIO presidencyApril 1941 Westbrook Pegler wins Pulitzer for

racketeering coverageSeptember 1941 - December 1941 Captive Mines DisputeDecember 1941 US enters Second World WarDecember 1942 Lewis Announces Drive to Organize DairiesJune 1945-June 1947 Reconversion strikesJune 1945 Ball-Burton-Hatch Bill introduced in SenateSeptember 1945 CIO launches Operation DixieMarch 1947 Taft-Hartley introduced in HouseMarch 1947 Utah adopts its ’little Taft-Hartley’June 1947 Taft-Hartley veto override

32

Chapter 2

What Did the Public Read About

Strikes?

Introduction

As emphasized in the previous chapter, the states adopted similar policies hav-

ing followed similar trajectories amidst heterogeneity. A key historical challenge aris-

ing from this observation is well-encapsulated thusly: ”What we have to study to

understand history is how structural forces cause people to change their notions of

what kind of situation they are in, and to sustain those new notions su�ciently long

to build them into institutions that in turn sustain them” (Stinchcombe 1978, 117).

The media are central to this process of structuration, for the popular notions of the

disruptive militance were drawn from the media, more likely than personal experience

(Gilens 1999, for example). Therefore, I begin by asking what did people learn about

strikes from the mass media, chiefly newspapers? Did these articles tend to foster a

pro- or anti-union outlook among citizens?

33

A potential worry confronting the researcher is that ”the concept of the media is

too vaguely defined to be useful in research, given the degree of change over time in the

composition of mass media” (Woolley 2000, 161). This is not unreasonable, but what

is at stake in the present study is the role of newspapers - as the predominant form of

news in this era - over a defined period of time. We are not evaluating a counterfactual

regarding whether the formation of labor law would have turned out di↵erently had

the newspapers and its leading columnist1 taken a di↵erent tack; rather, we are

evaluating the prospects for the labor movement to have avoided retrenchment.

The overall lesson of this chapter is that the framing of union militance and

the indexing of o�cial sources combined to buttress mass intolerance toward worker

militancy. Baumgartner and Jones (1993, 249) point out that ”in American politics,

issues that are phrased in terms of class interests, or as protecting the lower classes

against business interests, have little chance of success.” Since society could not exist

in the absence of labor relations, the cessation of labor production implied a direct

threat to the social order, or equilibrium. Though the salience of strike coverage was

diminished by war coverage, this does not imply that the prevailing frames were less

relevant (Ibid 1993).

Paletz and Entman (cf Sexton 1991, 270) argue that the ”general impact of

the mass media is to socialize people into accepting the legitimacy of their country’s

political system [...] direct their opinions in ways which do not undermine and of-

ten support the domestic and foreign objectives of elites [...] and deter them from

active meaningful participation in politics - rendering them quiescent before the pow-

erful.”

1Pace Lippmann. Though he was a highly regarded and longer-running syndicated columnistthan Pegler, Lippmann did not occupy a partisan niche in contrast with Pegler, who was firmly onthe right.

34

If the history of United States labor relations was a kind of private warfare,

then the recognition contracts won by the young CIO in rubber, auto, textiles, and

steel, were a kind of ceasefire. And the purpose of that ceasefire was for newly

defensive capital to reorganize and reverse labor’s programmatic gains. As I wrote

earlier, the unions gained whatever they won by means of the strike. Because the core

industries were closely coupled, the gains for labor in one industry implied further

gains (or losses) in others. Victories and losses in steel and coal were closely linked

with the industries of auto with rubber and steel (Fine 1969; Goldfield 1989; Zieger

1995).

Consonant with the previous chapter, I will analyze the volume as well as the

content of newspaper coverage of post-Wagner episodes of protracted and salient

labor disputes. The two central questions of this chapter are: how much were citizens

exposed to news content on strikes and the CIO; and whether there was a pro-con

directional thrust to that news. I proceed in four parts below: first, I review the

theoretical expectations with respect to news content and exposure assumptions; then

I turn to the data and method; then, I discuss the results, and I conclude in a final

section, connecting the lessons of this chapter to the study of public opinion in the

next.

I will proceed by analyzing the volume and content of news and columns about

strikes. This chapter begins its analysis by examining the count of articles on strikes

in major newspapers against the volume of strikes during the Wagner to Taft-Hartley

period. Then it turns to the content of selected articles to investigate how worker

protest and organization was constructed in the popular imagination. I investigate

coverage content in five ways: first, I examine the volume of strikes as covered in the

press over time; second, I present counts of the volume of coverage of labor leaders;

35

third, I examine the volume of key descriptions of the motivations of the strikers;

fourth, I study the content of coverage of key instances of salient CIO militance; fifth,

I examine the columns of the then-popular syndicated columnist, Westbrook Pegler.

Syndicated columns became increasingly popular during the 1930s and they facilitated

homogeneous messaging across the nation. Foremost among these columnists was

Westbrook Pegler, the 1941 Pulitzer Prize laureate for journalism2 wrote a column

that ”went out six days a week to 174 newspapers that reached an estimated 10

million subscribers” (Witwer 2005, 2).

It will be graphically apparent that the coverage of labor became less positive

and less union sourced. This negative tenor in coverage was notably linked to the

increase in denunciatory statements by Roosevelt administration o�cials. In the Flint

episode, I collect more than 250 articles and for the captive mines episode, I collect

150 articles; the other episodes had received rather less front-page coverage.

However, in order to prevent an inferential bias arising from beginning the

analysis after the CIO was instantiated, I take a look at the articles on the disruptive

strikes that marked the decline of Norris-LaGuardia and accompanied the adoption

of the Wagner Act - the 1935 textile and rubber recognition strikes. Furthermore,

the newspaper coverage of these strikes serves as a baseline for the analysis of strikes

later in the New Deal. Importantly, the strikes that precede the Wagner Act were

essential and prior to the formation of the CIO.

Theory and Expectations

In his seminal work, Gans (1979, 40-1) identified ’enduring values’ through

which journalists construct narratives and understand events; chief among those

2The first columnist to win the award, in fact. Though in some respects the forerunner of right-wing ’infotainers,’ his work during this 1930s and early 40s was widely respected.

36

frames being ’individualism’ and ’moderatism:’ ”one of the most important enduring

news values is the preservation of the freedom of the individual against the encroach-

ments of nation and society [...] the idealization of the individual could result in

praise for the rebel and the deviant, but this possibility is neutralized by an enduring

value that discourages excess or extremism. Individualism which violates the law, the

dominant mores, and enduring values is suspect; equally important, what is valued

in individuals is discouraged in groups.” That is, individuals might be hailed for tak-

ing visible stances outside the mainstream, but groups are unlikely to be lauded for

analogous stances.

These values are in opposition to the deployment of disruptive labor stoppages.

Iinterpreting strikes through the lens of moderatism and individualism paints strikes

and strikers as rebellious and highlights the costs of idleness as against the costs of

shop floor depredations; moreover, anti-militance messages and sources are privileged

over pro-militance ones. This is resultant from the phenomenon of news indexing

(Bennett 1990), meaning that the news organizations weigh their sources and frames

according to a power index: who is relatively more powerful in the situation being

reported. In the case of labor militance, the empirical question that arises is whether

union sources or ownership sources are being relied upon and in what proportion.

And, perhaps more important, it is a matter of whether public o�cials are delivering

pro or con messages with respect to labor.

In order to investigate whether mass opinion might have been manipulated, I

collect articles from the various outlets that most citizens relied upon. The condition

of manipulation cannot obtain unless the editors and owners intentionally presented

biased or selective information with the purpose of misleading readers (cf. Page

and Shapiro 1992, 356). This condition of intentionality is di�cult to demonstrate;

37

indeed, it is di�cult to demonstrate that the information presented was itself slanted.

At all events, the public did come to support policies that were inimical to their

interests. I proceed under the assumption that citizens shared a common framework

for understanding the disruptive strikes. Primarily, there was a shared valuation of

civil order. Or, at a minimum, there was a shared evaluation of what constituted

disorder.

How does a news article a↵ect the reader? First, citizens must be exposed to the

article(s). The canonical political science understanding is that political awareness is

the best predictor of receiving news messages (Price and Zaller 1991). Unfortunately,

these early Gallup surveys do not include political awareness items. For this reason,

the subsequent analysis will focus on front page news so as to make my exposure

assumption minimally heroic. If an article has an e↵ect across people in a uniform -

or at least, similar - fashion, then it must be interpreted and received in a uniform,

or at least similar, fashion. In part, processing is a function of the political values

elicited by the content. In their resistance, the CIO unions threw the patterns of

social and economic relations into deep question, even property itself (Lichtenstein

1995). The capitalists (and the government) in their repression of strikers exposed

the limits of revolutionary resolve on the part of the militant unionists.

The unionists were portrayed as a vector of social disorder by the various outlets;

and, not unreasonably, the public came to see unions as an interest group. This was in

part because of the personality and style of the leader of the CIO John L. Lewis. Some

discussion of where he fits is in order here, as his name was a metonym for organized

labor militance (Zieger 1988, Chapter 4). His opponents within and outside the labor

movement used his connection to the communist militants in the CIO to try and

turn mass opinion against him. Ultimately, however, the depiction of unionists as

38

separate interest from the commonweal was likely a partial result of Lewis being a

fairly self-interested labor leader. To wit, Lewis was a proponent of essentially laissez

faire capitalism in the 1920s (his argument for union security legislation was that it

would fix labor costs among the various competitors so as to level competitive pressure

within the coal industry). Moreover, Lewis’ proximate interest for the formation of

the CIO was to bring recalcitrant steel owners to heel (James Wechsler Papers, Box

1, Folder 2).

President Roosevelt expressed concern about the syndicated columns and the

ways in which they seemed to be shaping opinion against him (White 1979, 27). In

Graham White’s seminal study, he found that the press was in fact not particularly

biased against the president (Ibid, Chapters 2-4). However, following his approach,

in that I juxtapose coverage of identical events, I might weigh the relative di↵erences

in sources and directional thrust of statements.

Also, I pay attention to the rise of the use of the word ’management’ in the

place of ’capital.’ After all, in any reasonable formulation, the antonym of capital

is labor. The concept of management as a distinct entity accords the workplace a

special location as a site of scientific control. This in turn reifies the ’public interest’

as an outcome that is coterminous with industrial and shopfloor order. The scientific

notion of management implies that industrial disputes have a fundamental solution

that is evidently in the public interest. This concept of the public as a third party to

industrial disputes was one shared by the workers themselves (Harris 1982; Rondinone

2010, 147-50). That is, the workers did not conceive of their e↵orts - even in the

disruptive sit-own strikes - as a class force let alone fighting for industrial democracy;

they were fighting for shopfloor improvements. In a nutshell, management coming to

stand in for capital served to mystify economic relations. Then, management came

39

to be linked in the public imagination with smooth production of materiel in a just

and successful war e↵ort. The upshot was a national perspective that was, at its core,

hostile to labor militance.

What these selections will illustrate - and all these are articles in the content

analyses were on the first page of the papers - is that readers were generally exposed to

frames of the sit-down strikes as sites of disorder, as opposed to sites of well-motivated

protest. In the previous section, I suggested that the news media was in concert

with the capitalist elites. But this is a di�cult proposition to sustain, particularly

given that editors do not ”directly take the audience into account when selecting

and producing stories but instead assume what interested them would interest the

audience” (Gans 1979, 229-30). That is, the fact that the newspapers produced and

framed stories against the industrial unions is not enough to press the case that they

were in league against them. One might well have concluded that the disorder frame

was the best way to sell papers.

The journalist need not be ideologically in line with the capitalist owners in

order to slant the news. Nor need the ownership of the papers be involved on a daily

basis in order to maintain a pro-capitalist slant in terms of new or editorial slant.

The reporter may engage in self-censorship (Gans 1979, 196) or the editors may

discourage or pressure their subordinates. The tension between the profit motive

and the First Amendment responsibilities of the press is laid bare in the coverage of

strikes. During the episodes we will examine below, the media did not challenge the

anti-labor interests of media owners or capital-holders generally.

With respect to the interest of newspaper editors, the rise of syndicated columns

served their end of promoting the American tradition of union suppression. Earlier,

I identified Pegler as a subject deserving of focus because of the acclaim he received

40

during the 1930s and 40s and, more importantly, because the widely disseminated

Scripps chain featured his columns across the nations. Pegler’s reportage on union

racketeering in Northern cities exposed notable corruption. Pegler was among the

first to use the term ’right to work’ and apply it to the right for workers to break a

picket line - the more intuitive reading for many is as a kind of label for a full em-

ployment program. ”This [the Wagner Act] is a law which drives millions into unions

against their will, under the rule of unioneers [sic] whose character the government dis-

claims any right or responsibility to investigate” (Washington Post 10/15/39). But

the reactionaries were not altogether unreasonable in their branding because what

was putatively at stake was citizen liberty to work for a wage in order to make a

living.

Data and Method

The focus of this chapter will be twofold: how much news were people exposed to

and what was the valence of that information. First in order to measure the overall

volume of articles relating to strikes, I begin by searching the ProQuest Historical

Newspapers database of twenty newspapers in large cities for articles with ’labor’

or ’union’ and ’strike’ mentioned in both the headline and main text of the article.

This yielded a total 54,633 articles from 1936 to 1947. The count of strike articles

comprised a large proportion of the total articles regarding union activity, which was

100,978. About a sixth of the articles are drawn from the New York Times and New

York Herald Tribune; the remaining proportion are roughly equally divided among

the other big city newspapers (i.e. Boston, Washington, Chicago, Atlanta), and then

the remaining newspapers comprise a small proportion of the overall count. Broadly

speaking, the overall coverage of unions was dependent on their militance; that is,

the coverage of unions was visibly related to strikes. (See Appendix for outlets and

41

search strings.) This database relies on having archived long-running big-city US

newspapers. In addition, I include newspapers targeted at the black community,

though these contained only a few instances of strike coverage.

With respect to the content analysis, like Page and Shapiro (1983; 1987), my

independent variable is pro-con statements made by valenced sources. Each mea-

surement, then, is a unit of analysis in measuring aggregate newspaper content. The

content analysis in this chapter will be comprised of 362 articles spanning five key

episodes in the New Deal period. Though this study relies on a single-coder, an audit

of the items in the Appendix will reveal that the coverage of unions was unfavor-

able. However, demonstrating that the coverage was unfavorable does not in itself

demonstrate that the coverage was unfair. Indeed, a reader might well view that the

instances in this chapter’s Appendix as instances of factual reporting. However, the

way that these reports stack up over time and across contexts, doubtless redounded

to labor’s detriment. There are many ways to understand a strike but the newspa-

pers preferred to understand the strikes as being in opposition to social harmony or

patriotic demands.

In this chapter, I measure pro-con source-messages in several contexts: most

comprehensively, I measure the directional thrust of coverage by analyzing the content

of coverage of key episodes during the Wagner to Taft-Hartley interlude. In the

Appendix, the reader can find each instance of an anti-CIO or anti-worker protest

item cited in the figures. The strikes were, not altogether unfairly, implicitly described

as insurgent in many cases in most newspapers. The main expectation is that a lens

sympathetic to the cause of worker resistance would be relatively uncommon and

increasingly so during the Wagner to Taft-Hartley period.

The data are summarized in tables that depict the overall thrust of the source-

42

messages in the articles wherein articles are classed as favorable, unfavorable, bal-

anced, or neutral with respect to strike actions. A favorable item was one where

the balance of source-messages was supportive of labor; an unfavorable item was one

where the balance of source-messages was opposed to labor; a balanced item was one

where the source-messages were even; and, finally, a neutral item was one where there

was no directional thrust to the source-messages.

Woolley (2000, 165-6) advises the use of a deflator to adjust for the implicit

selection biases arising from the indexing process. Here ’indexing’ connotes the pro-

cess by which the newspaper archive arranges and classifies articles. These biases are

unavoidable and can only be partially remedied by evaluating the search results by

hand. Therefore, I might report counts relative to the total number of articles about

strikes indexed when dealing with the issue of attention paid to unions. But, the

visual juxtaposition of the article counts in Figure 1 below suggests that I need not

report the counts as ratios of the total number of indexed articles because both counts

move in tandem (r=.93). At all events, this concern is orthogonal to the matter of

news exposure.

To justify the focus on Lewis, I compare Lewis mentions to the those of other

prominent labor leaders. Lewis dwarfs the other principals of the CIO in news cover-

age. Only the radical communist Harry Bridges of the Longshoremen’s union received

nearly as much attention as John L. Lewis in any given year. This is consistent with

the view that the militant workers were being brought under a disproportionate media

spotlight. Lewis coverage was especially disproportionate to that of the contempo-

raneous head of the AFL, William Green whose newspaper count was a fraction of

Lewis’. This imbalance, I argue, is a manifestation of a long-standing media bias to-

ward conflict, following canonical works in the social science literature on the media

43

(Gans 1979; Iyengar and Kinder 1987, for example). Conceptually, the methodolog-

ical problem of index bias does not arise when I deal with the media attention to

labor leaders because I am comparing attention paid to labor leaders merely to es-

tablish Lewis’ primacy among them, so any articles that were excluded by those who

arranged the database should a↵ect the various union principals equally (or at least

randomly). As a contemporaneous and long-time Lewis observer, James Wechsler,

pointed out,

Admittedly, [the] CIO was the creation of many anonymous men, of industrial pres-sures, of forces that might not have been indefinitely suppressed. But Lewis was itsfounding father, its emotional symbol, its commander-in-chief. And nearly all of therollicking recruits who signed CIO card identified the movement with the name andperson of John L. Lewis. So did Mr. Lewis. (James A Wechsler Papers, Box 1, Folder2)

In order to identify pro-labor or anti-labor source-messages through content

analysis in a replicable fashion, despite the interpretive nature of the exercise, I pro-

ceed by analyzing coverage of identical episodes by di↵erent outlets. I focus on several

salient episodes guided by insight from the history literature: the Flint sit-down strikes

of 1936-7; Lewis versus Hillman in the breakup of the original CIO in 1937-403; the

captive mines4 dispute of 1941; and the Reconversion strikes of 1945-6 (Fine 1969;

Zieger 1988; Lichtenstein 1982[2003], for example). I code an source-message as ’con’

when it contains references to militance that cast aspersion upon the workers or sets

the struggle as inherently in opposition to civil society, delivered by a source from

ownership or a public o�cial. And, mutatis mutandis, I code a source-message as

’pro’ when the strike is contextualized within existing grievances or is delivered by

3Lewis resigned the presidency of the CIO in after the 1940 convention because the unionists inleadership and rank-and-file supported FDR instead of demanding more concessions.

4By ’captive,’ it is meant that the coal mines were vertically integrated with the steel firms thatused their output. Thus, the mines were e↵ectively captured by the steel firms.

44

a union source. The typical negative framings of worker militance during this pe-

riod held that strikes were to be abhorred because they excluded men who wanted

to work; because they were in opposition to property and, by extension, the rule of

law; and because the news tended to index the views of o�cial sources, the source-

messages in question are pro-capitalist or pro-worker in that they reported events in

a slanted fashion, chiefly by editorializing in news sections. Again, the limitations of

single-coder analysis notwithstanding, it will be evident that the coverage of unions

cannot have been construed to have been generally favorable toward unions, even if

a skeptical reader would argue the coverage was fair.

Also I will examine the columns of Westbrook Pegler, who arguably foreshad-

owed the latter-day syndicated conservative columnists; Pegler’s columns, usually

featured above the fold across from editorials, formed the basis for a newly crystalliz-

ing American conservative perspective on labor unions that emphasized their corrupt

elements and the costs borne by workers a↵ronted by the picket line. The construction

of unions as corrupt and as obstacles to an orderly society find some of their earliest

wide expression in Pegler’s columns (Kennedy 1999; Witwer 2005). A major reason

that Pegler was granted national reach was that Roy Howard (of Scripps-Howard) had

turned against the New Deal in 1937; moreover, Howard sought to ape the national

cohesion of the Hearst papers (White 1979, 190; Emery 1972, 431).

Analysis

Figure 2.1 depicts the volume of front-page coverage of strikes plotted alongside

coverage of John L. Lewis, the CIO, racketeering, and communism. The count of

articles on strikes notably spiked in the major years of labor unrest of 1937 and

1946 and the volume of CIO coverage likewise spiked in those years. The John L.

Lewis volume varied with volume of strike articles for the most part (r=.73), as he

45

0

500

1000

1500

2000

2500

3000

1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947

Art

icle

Cou

nt

Year

John L. Lewis Articles

Racket Count

CIO Articles

Communism Articles

Strike Articles

Source: ProQuest Historical Newspapers

Figure 2.1: Strike Coverage Per Annum

was personally linked - negatively - with disruptive militance, even after his break

with the CIO. Though the coverage of communism in strikes was relatively muted, it

also tracked the overall coverage of strikes (r=.38). The racketeering frame, however

(r=.33), had little to do with the prevailing levels of militance and received relatively

little attention in the mainstream press, in contrast to its salience for right-wing

backlash to unionism in the 1930s. Overall, the sheer volume of strike coverage on

the front page suggests that citizens in the mass public were well-acquainted with

portrayals of labor militance. During the peak strike years, citizens were exposed to

more than 7,000 front page stories about strikes per year; in other words, at the height

of industrial conflict, the nation was blanketed by strike coverage. And, importantly,

strike coverage was at a peak in 1937, evidencing this issue’s salience in the subsequent

elections.

46

0

500

1000

1500

2000

2500

3000

3500

4000

0 500

1000 1500 2000 2500 3000 3500 4000 4500 5000

1936

19

37

1938

19

39

1940

19

41

1942

19

43

1944

19

45

1946

19

47

Str

ike

Volu

me

Art

icle

Cou

nt

Year

Strike Articles

CIO Articles

Strike Volume

Figure 2.2: Strike Coverage and Strike Volume

Figure 2.2 shows the count of strike articles plotted alongside a strike volume

measure (cf. Hibbs 1976) and the union approval measure from the next chapter, re-

spectively. In Figure 2, we observe that the coverage of strikes and the strike volume

vary together until the Second World War when the coverage of strikes increased as

the strike volume was flat. During the immediate post-war period, citizens ranked

industrial disputes as the most important problem facing the nation. Though Gallup

did not ask the question in a su�cient number of instance to analyze along the lines

of Funkhouser (1973), this result does suggest that newspapers inflated the strike

problem. We will see in the next chapter that movements in union approval moved

inversely with the volume of strike coverage; as the volume of coverage increased the

overall level of union approval tended to decline.The volume of strike coverage was

clearly linked with the strike volume measure (r=.57) borrowed from Hibbs.5 Further-

more, the count of strike articles has an even stronger relation with the count of CIO

articles (r=.89). The count of articles was negatively, though insignificantly, related

5See Chapter Four.

47

with the expressed popular approval of unions (r=-.04) and strongly negatively corre-

lated with approval of John L. Lewis (r=-.51).6 Relatively more anti-strike coverage

emerged over the course of the sit-down strikes as shown in Figure 4 below. And,

indeed, the framing of the issue is suggestive in support of capitalist power. As noted

above, the negative coverage in newspapers took two forms: first in the descriptions

of union militance and second in the sourcing of reportage. Taken together, these

results are consistent with the hypothesis that the valence of coverage buttressed

popular disapproval of labor militance.

0

500

1000

1500

2000

2500

1936

19

37 19

38 19

39 19

40 19

41 19

42 19

43 19

44 19

45 19

46 19

47

Art

icle

Cou

nt

Year

Murray Count

Hillman Count

Lewis Count

Figure 2.3: Lewis, Hillman, and Murray Coverage

Figures 2.3 and 2.4 show a comparison of the volume of coverage of prominent

labor leaders. We can see clearly that the coverage of John Lewis greatly exceeded

that of any other labor leader. In presenting the labor leaders in this fashion, the

media was tending to highlight the most combative labor leader. In juxtaposing the

6I do not present the Lewis approval measure in the figures because the time-series is incomplete.

48

Table 2.1: Flint Coverage

NYT Herald Trib Balt Sun Wash Post CSM Bos Globe TotalFav. 20% 12% 22% 25% 26% 35% 13% 22%

Unfav. 46% 31% 30% 17% 19% 15% 26% 27%Bal. 32% 35% 41% 54% 55% 50% 52% 45%Neu. 2% 22% 7% 4% 0% 0% 9% 7%N 40 26 27 35 31 20 23 221

coverage of the labor leaders, the media were e↵ectively presenting the choices facing

organized labor and the public over increasing militance and heightening demands as

opposed to subordinating to membership in the FDR coalition. And, indeed, Lewis

was cast as out of the mainstream by much of the press. Though Lewis was not a

metonym for labor, he was by far the most famous labor leader.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

1936

1937

1938

1939

1940

1941

1942

1943

1944

1945

1946

1947

Art

icle

Cou

nt

Year

Green Count

Bridges Count

Lewis Count

Figure 2.4: Lewis, Green, and Bridges Coverage

Table 2.1 depicts the measures for the coverage of the Flint sit-down strike.

A total of 221 articles were recovered from the search. In these items, most were

49

Table 2.2: Hillman and Lewis Coverage

ProportionFavorable 14%

Unfavorable 29%Balanced 43%Neutral 14%

N 14

balanced or neutral, though there were more unfavorable items than favorable. The

New York newspapers were more anti-strike than those elsewhere. It should be noted

that numerous strike articles were written by the wire reporters and therefore many

of these items were identical across the nation.

Table 2.2 deals with the juxtaposition of John L. Lewis and Sidney Hillman in

the newspapers from 1937 to 1940, the final year the original CIO would be intact.

Here the papers are not disaggregated because there were relatively few articles con-

trasting the two During the run-up to FDR’s second reelection, Lewis argued that

the industrial unions should have opposed Roosevelt in the absence of specific con-

cessions to organized labor. Hillman, saw that labor needed the president more than

he needed the upstart unions. This logic is based on the idea that the workers could

not carry on using their collective might unilaterally to impose political and policy

demands. Lewis, for his part, in trying to utilized voluntarist logic was missing out

on how hostile the Republican Party was. The only viable plan for Lewis, which he

came upon after his resigning as head of the CIO was to seek a farmer-labor alliance.

In this table, the balance of the articles is again mostly that of general discussion

an dispassionate reportage; however, to the extent that a slant is noticeable, it tends

toward the anti-Lewis side.

Table 2.3 depicts the coverage of the 1941 captive mines dispute. These strikes

50

Table 2.3: Captive Mines Coverage

NYT Herald Trib Balt Sun Wash Post CSM Bos Globe TotalFav. 26% 13% 25% 15% 19% 15% 0% 15%

Unfav. 55% 46% 75% 70% 43% 39% 50% 51%Bal. 19% 37% 0% 15% 38% 39% 33% 28%Neu. 0% 4% 0% 0% 0% 7% 17% 6%N 31 22 8 13 21 13 12 150

Table 2.4: Reconversion Coverage

ProportionFavorable 18%

Unfavorable 32%Balanced 47%Neutral 3%

N 39

were an attempt to unionize the mines that were owned and operated by steel firms for

use in production. The National Defense Mediation Board - a tripartite commission

- ruled against the United Mine Workers (Atleson 1998, 20-31). In response, Lewis

called a headline-grabbing strike that would have served as a grave obstacle to war

production (Zieger 1988, Chapter 4). (Public disfavor in survey responses regarding

these selfsame strikes are presented in the next chapter.) A total of 150 articles were

found in this search. This table shows the coverage of the captive mines dispute of

1941, which occurred during the run-up to World War Two. In this episode, the

negative coverage was spread across the national outlets. And, again, only a small

proportion of the articles might be said to be have treated the militant unionists

favorably. The majority of items are coded as unfavorable in this instance. In contrast

with the Flint episode, the captive mines episode had more negative coverage than

neutral coverage. The key di↵erence in this episode being that the union militance

was places in direct opposition to the war footing; this is in keeping with the highly

favorable view of defense strike bans examined in the next chapter.

51

Table 2.5: Pre-Wagner Coverage

ProportionFavorable 33%

Unfavorable 11%Balanced 53%Neutral 3%

N 33

Table 2.4 depicts the coverage of the reconversion strikes. These connote the

strikes that attended the ’reconversion’ of the economy from a wartime to a peacetime

basis during 1945-6. The Reconversion strikes, which were more numerous than the

size of the sample studied here would suggest, were more largely sourced o�cially than

previous episodes. Nonetheless, the Truman administration was not as denunciatory

as the preceding administration. The N here is only thirty-nine, but this is because I

am selecting strictly for the articles that frame the strikes as reconversion strikes. The

Reconversion strikes, chiefly against General Motors and US Steel, were designed to

insert labor into the post-war management paradigm. Materiel loss was a key theme

though out the wartime period, but the dominant negative frame here was that of

insurrection.

Table 2.5 depicts the coverage of the disruptive wave of textile and rubber strikes

that helped form the CIO in early 1935 (cf. Bernstein 1966; Zieger 1995). A search

of strike articles related to textiles or rubber on the front page in 1935 before the

adoption of the Wagner Act yields 33 results. The pre-Wagner coverage, though, was

not ideologically more favorable to unions. And, as previously, the negative mentions

are presented in the Appendix. Here, though, 33 percent of the articles were favorable

compared to 11 percent unfavorable. In this coverage, the e↵orts by the President and

senior o�cials to forestall and broker labor disputes was favorably covered. In other

words, there existed a manifest bias toward labor peace even in otherwise neutral

52

coverage.

By way of example, here are quotations from earlier in the sit-down wave.

From the New York Herald Tribune, ”Ten thousand persons were thrown out of work

today when the B.F. Goodrich Company suspended production because of a series

of departmental ’sit-downs” (9/24/36). From the Baltimore Sun of the same day

”Labor disputants reported to fists and stones in two widely separated cities today,

while others sought to end strikes though conciliation.”

In order to give the reader exemplars of what is meant by a ’negative’ phrase7

in the lead paragraph, I will present randomly selected articles coded ’negative.’

From the February 4, 1937 New York Times, explaining why the Sheri↵ of Flint,

Michigan called for the National Guard to encircle the ”thousands of sitdowners

and pickets there that had defied him to compel its evacuation and end picketing

under the injunction issued” (New York Times 2/4/37). From the February 25,

1937 Los Angeles Times on page three, ”A bitter clash of opposing union leaders

yesterday marked mediation of a paralyzing sit-down strike at the giant Douglas

Aircraft Corporation” (Los Angeles Times 2/25/37). From page one of the March 22,

1937 Chicago Tribune, ”Professional agitators with communistic sympathies, carefully

trained for their duties in a New York City labor college, are the storm troopers

and the backbone of the present strike wave’ (Chicago Tribune 3/22/37). From the

Washington Post of the same day, ”It is surprising that President Roosevelt with

his proper concern for the many grave problems which confront the Nation, has as

yet found no occasion to express himself publicly on the subject of sit-down strikes.

The crisis in Detroit where a great city is rapidly drifting into a condition not far

7I do not report positive mentions because relatively few existed. Though there were fairlypositive features in newsmagazines such as Harper’s and Life, newspapers were either factual intheir introductory paragraphs or biased against the industrial unions.

53

removed from anarchy” (Washington Post 3/22/37). The antagonism over the sit-

down strikers was so pitted that the Governor Frank Murphy (D-MI) was the target

of death threats as ”some of those threatening him assail him for not adopting more

vigorous politics in the ejection of sit-down strikers, while others would have revenged

themselves on him for permitting such ejections” (Chicago Tribune 3/22/37).

In these articles, the strikes are often described using a metaphor of armed

conflict and battle. John L. Lewis was referred to or likened to a leader in combat on

a consistent basis. It is di�cult to identify negative mentions of Lewis as opposed to

the strikers or unions generally. Adjectives such as ’unyielding’ - commonly applied to

Lewis - are more ambiguously valenced when applied to individuals than to interest-

groups. A typical report on Lewis: ”John L. Lewis tonight exposed what he considers

his ace in the hole in the automobile strike. ”We helped Roosevelt lick industry, now

let Roosevelt help us lick industry” (Chicago Tribune 1/21/37). This invocation of

Roosevelt here implied that in exchange for CIO support in reelection, the President

would support the sit-down strikers.

As Figure 2.5 shows, newspapers carried twice as many articles about labor

union demands than about labor union grievances or pleas. The distinction be-

tween ’demands’ and ’pleas’ is found in the implicit power-relation connoted by the

terms: where a demand is imperious, a plea is suppliant. The count of the latter in

this analysis is somewhat inflated by articles concerning grievance machinery within

firms. Pairwise correlation with union approval in survey responses shows a positive

correlation for ’plea’ (r=.32, N=12) and a small negative correlation for ’demand’

(r=-.16, N=12) from 1937-48. These correlations imply that unions were more likely

to viewed favorably when they were framed as supplicants rather than as insurgents.

And, indeed, there were 10,221 articles citing the stated preferences of unionists as

54

0%

10%

20%

30%

40%

50%

60%

70%

80%

0

200

400

600

800

1000

1200

1400

1935

19

36

1937

19

38

1939

19

40

1941

19

42

1943

19

44

1945

19

46

1947

19

48

% A

ppto

ving

Uni

ons

Art

icle

Cou

nt

Year

Plea Articles

Demand Articles

Union Approval

Source: Gallup/AIPO; ProQuest Historical Nespapers

Figure 2.5: Use of Plea v. Demand

’demands’ and 4,072 referring to ’pleas.’

Rather than referring to capital as such, ’management’ began to prevail as the

preferred moniker. And, more subtly, the description of labor comes to be synony-

mous with unions and not with the class itself (Rondinone 2008, Chap 6). In other

words, there was an absence of apprehension of class conflict during the pitted mili-

tance. The idea of management - a sub-class of capitalists, strictly speaking - serves

to deny the urgency of the claims and demands of either side and consigns the conflict

into a technocratic realm. And a scientific outlook on management implies an incor-

rectly pluralistic take on American labor relations (Harris 1982, 7-9; Atleson 1983).

Necessarily, the government comes to be viewed as a player in strikes alongside capital

and labor.8 Interestingly, Governors Murphy of Michigan and Earle of Pennsylvania

8This is a distinct point from saying that all politics constitutes the ordering and sublimation of

55

were often protagonists in the articles about the strikes in their states. The rise of

the perceived legitimate role of government in sorting out labor militance was related

to the news media predilection for defusing class war.

Pegler helped make racketeering within unions a salient issue for many Amer-

icans. In terms of identifying unions as actors, he drew the unions as a gatekeeper

over jobs that free people seeking work. For example, ”The unioneer’s solemn as-

sertion of a man’s right to work is sacred until it violates the decree of a few union

politicians whose appetite for power has been tempted by a group which seems ready

for a raid” (Washington Post November 11, 1937, 11). Also, ”The fact is that the

labor leader very often is a dictator maintaining his power by union politics at the

expense of democratic procedure which places the rank and file in the position of

the Italian or German citizens who just rock along submitting to discipline” (The

Atlanta Constitution June 6, 1937, 4). The racket issue brings Michels’ iron law of

oligarchy into focus. The leaders who are trusted to ensure that the worker cartel is

administered and organized have distinct and sometimes opposing incentives to the

rank-and-file members. Senator Wagner, introducing his eponymous law, claimed ”I

do not underestimate racketeering in organized labor. It exists, unfortunately, as it

exists in politics, in industry, among lawyers, and with bankers and brokers” (NLRA

Legislative History 1935, 1319).

In March 1941, across classes, respondents overwhelmingly responded in the

a�rmative to the following item: ”Westbrook Pegler, the newspaper writer claims

that many union leaders are racketeers.”9 Do you agree or disagree with him?” (Gallup

conflict.

9’Racketeering’ connotes a fraudulent service - a service that would not exist were it not for theracket. In this case, unions, such as the Teamsters (Witwer 2004, for example) used their control ofjobs to compel incoming workers to surrender wages.

56

3/12/41). Survey results reflect the absence of the expected class cleavage on this

item. Those on relief and the wealthy overwhelmingly agreed with Pegler that unions

were largely rackets. Every region and class sub-group agreed at a rate greater than

70 percent Moreover, there was little regional variation, except, unsurprisingly, the

South was especially likely to disparage labor leaders. Astonishingly, a majority of

union members that responded answered in the a�rmative. It is typically a mistake

to dwell on one poll result but this response is starkly negative.

There is a relevant normative consideration here: that of the individual rights

of the workers wishing to cross the picket line. And here the researcher cannot avoid

taking a normative stance as the pluralist neutrality redounds to an a�rmation of

capital’s preeminence over labor. The choice is whether would-be free riders can be

delivered benefits on a particular and selective basis from group activity; since the

union’s gains are non-excludable in the workplace, the union is impelled to demand

dues. Pegler’s insistence on the right of workers to cross picket lines as well as for

the ’public’ not to be deprived of goods articulated the conservative backlash to

union power and served to anticipate the o↵ensive against unions that attended the

’right turn’ of the 1980s and the preset-day opposition to public sector unions. Put

another way, the discourses surrounding the rise of the New Deal were the ones that

surrounded its fall a half-century later.

In his column, cheekily entitled ’Fair Enough,’ Pegler wrote that ”organized la-

bor is assuming the powers of government without the responsibility of government,

and with business in shackles that is a lopsided arrangement which cannot continue

indefinitely” (Los Angeles Times, 8/8/39). By this, he meant that the use of strikes

that seized the site of production or obstructed the flow of commerce were an arroga-

tion of power by the unions. Worse yet, the unions were not necessarily doing their

57

workers bidding. Like Michels before him, Pegler had the insight that the incentives

faced by workers were distinct and potentially contrary to those faced by labor leaders

hungry for dues. Moreover, Pegler was critical of the perceived violence of strikers

who assaulted those who meant to cross picket lines. Basically - and this is more

true of the CIO unions - the nascent union movement of the 1930s was characterized

by an expansion of internal governance structure that has its own taxation scheme

and ability to annex parts of the United States. And, indeed, the GOP would make

curbing union power the central theme of the 1946 midterm elections, capturing the

spirit with the slogan ’Had Enough?” (Lee 1966, 47).

The relationship between capital and labor was antagonistic throughout US

history. And newspapers took di↵erent perspectives, though typically maintaining a

pro-capitalist bias. Unhappily, there do not exist valid measures of public opinion

prior to the adoption of the Wagner Act. Nevertheless, we can surmise that the

coverage of the strikes before the Wagner Act were not substantially more favorable,

and to the extent that the coverage of militancy became more negative over the

course of the New Deal, it appears attributable in large measure to the increasingly

negative stances taken by the Roosevelt administration and the president personally;

Roosevelt took his his strongest stance against militance during the captive mines

dispute.

A sampling of newspaper coverage - even a more comprehensive one than that

presented above - would be insu�cient to summarize the news that citizens received

about strikes, as radios became more commonly owned throughout the period under

study. However, newspapers were the dominant source of information. And, sub-

stantively, what they have to teach us would not be materially a↵ected by attention

to newsreels or radio, as it is unlikely that the newsreels and radio presenters would

58

have been delivering markedly di↵erent content.

Newspapers in Utah and Wisconsin

Continuing with the theme of analyzing states that had protective laws before

restrictive laws, I examined the front pages of newspapers in Utah and Wisconsin with

an eye toward stories about labor disputes. Given that the New York newspapers

comprise a substantial fraction of the national dailies in the analysis presented above,

this section will take a brief look at Utah and Wisconsin papers which were not dig-

itally archived in a searchable fashion. But there are numerous examples of negative

framings of strikes from the New York papers in the Appendix. Further, it should

be noted that the presses with narrow circulation often reprinted the big-city chains.

The generalized logic drawn from the big-city newspapers is applied to these states in

this section: here, we are looking for examples of strike coverage that are inconsistent

with the business interests of media owners. For Utah, I used the Deseret News, a

statewide daily that had the attention of Mormon Church members. For Wisconsin,

I used the Milwaukee Journal as this paper had the widest circulation in the state.

I attained these newspapers from the Google News digital archive. (Though that

archiving project was eventually abandoned, these collections are largely complete,

missing no more than a handful of articles per year.)

In contrast to the concerns raised about the Utah case in the previous chapter,

the First President of the Latter-Day Saints, Heber Grant, stated in a front-page

editorial that ”The Church does not interfere nor has any intention of interfering

with the fullest and freest exercise of the political franchise by Church members”

(Deseret News 11/2/36). One might not treat this statement in a credulous fashion

but the forceful denial itself evidences the importance the Mormons put on presenting

themselves as endorsing individual political freedom.

59

In covering strikes in silver and copper mines, the Deseret News distinctively

led articles on the strike by presenting the workers’ position rather favorably, for

example: ”the general rise in metal prices and commodities are reasons for the mine

operators to look with a friendly eye on worker demands” (Ibid 11/7/36, 1). However,

on the very same page, the paper presented strikes occurring in the San Francisco

shipyards as ”threatening paralysis for national shipping” (Ibid). And furthermore

that article began by pointing out that ”eggs were hatching aboard a strikebound

vessel” (Ibid).

In covering out-of-state strikes, the News was sometimes less than even-handed.

For example, ”Strikers attacked o�cers with stones when attempts were made today

to open a strike-bound cannery at Stockton” (Ibid 4/23/37, 1). But, on the other

hand, it o↵ered a fair contextualizing of a CIO strike against Jones and Laughlin

”Police hurled tear gas into a crowd of shouting, demonstrating pickets today at the

strike-closed Aliquippa works of the Jones and Laughlin corporation” (Ibid, 5/13/37,

1). Foreshadowing an issue this dissertation revisits in the conclusion, ”Striking steel

workers adopted a ’quit work or starve’ policy against men remaining in besieged

mills, and the Republic Steel Corporation answered with a protest” (Ibid, 5/29/37,

1). And this was not the only time the News relied upon a martial metaphor to cast

labor in an unfavorable light, referring to the ”strike-sieged plants of Youngstown”

(Ibid 6/28/37, 1). It should be noted, though, that this newspaper did spend more

time on the events in Europe than on industrial strife at home.

Overall, the Deseret News o↵ered a more even-handed approach than other

newspapers. Indeed, when Senator Elbert Thomas (D-UT) questioned a Repub-

lic Steel executive on the provision of weapons for Chicago police o�cers acting as

strikebreakers, it made front-page coverage (7/3/37) whereas the New York Times

60

and even Chicago Tribune relegated the story to well inside the paper. And when

Lewis began to turn on Roosevelt in 1937 the Salt Lake newspaper elected to em-

phasize the Labor Day celebrations at the expense of highlighting the growing rift

within and between labor and the administration (Ibid, 9/6/37, 1). Though the paper

hardly dealt with strikes on the basis of their national salience - indeed the Deseret

News did not report on the Flint sit-down as front-page news until nearly the end

of the strike. And, indeed, upon the passage of the state’s little Wagner Act, there

was no report of the bill’s passage whatever. But, by way of comparison with the

state’s other major daily The Salt Lake Tribune, the copycat sit-down strikes that

came shortly after Flint - which, after all, convinced General Motors to recognize the

CIO union - were described as an ”epidemic” on the Tribune front page while it did

not get mentioned in the News (Salt Lake Tribune 3/3/37, 1). And, what is more,

the News did not cover the captive mines dispute either. During the post-war strike

wave, the News was quick to point out that the steel and coal strikes would have little

e↵ect on Utah (Deseret News 1/16/46,1). However, the position of the editors of the

News should not be mistaken for progressive: indeed, the paper printed editorials in

opposition to child labor laws on the grounds that the ”state should be kept free from

federal usurpation” (Ibid 1/26/35, 12). Furthermore, they argued against supporting

Roosevelt’s reelection bid suggesting that, in light of numerous New Deal laws being

held unconstitutional, ”Church members must stand for the Constitution and should

feel duty bound to vote to protect it” (Ibid 10/31/36).

Upon the passage of the Wisconsin Labor Relations Act (WLRA), the Milwau-

kee Journal wrote ”On the hundred and first roll call of a continuous 12-hour session,

the hungry and weary state senate passed the administration’s labor disputes bill

in just the form its sponsors wanted it” (4/3/37, 1). On that same front page, the

61

Journal prominently features a quotation by Senator Borah (R-ID) that anticipates

a discussion that we will save for the final chapter: ”I confess I have not been excited

over the fact that labor in its struggle with capital has seized property but you cannot

separate this sole issue from all the others involving individual rights” (Ibid).

In the wake of the WLRA having been passed, the unions sought to organize

the creameries and dairies of Wisconsin. Examples of negative agricultural responses

were salient and featured in the Journal. Early after the passage of the Wisconsin

little Wagner Act, the Journal began an article about a strike as ”paralyzing the

produce trade” (Ibid 5/6/37, 1). Another example, ”Unionization of the Richland

Cooperative Creamery Company at Richland Center was protested Monday night

by 325 farmer-patrons” (Ibid 4/19/38, 1). The next month: ”Sixteen square jawed

farmers representing the Wisconsin Dairy Industries Cooperative declared here Mon-

day that they are prepared to don ’political brass knuckles’ for a showdown resisting

unionization of farmer-owned plants” (Ibid 5/17/38, 1). And the next day, ”Oper-

ators of twelve large Milwaukee poultry and egg firms took the initiative in a labor

controversy and closed their commission houses after negotiations broke down over a

closed shop provision” (Ibid, 5/18/38,1). The Journal featured the farmers’ backlash

prominently, ”Union Grove farmers who were ’innocent bystanders’ of the spinach

strike last spring started to do something about their plight Monday. ’No closed shop

on the farm!’ was their keynote” (Ibid 7/26/38,1).

Having won a sit-down strike against Allis-Chalmers in May 1938, the Journal

framed its front-page coverage of the union victory thusly, ”Immediately upon [col-

lective agreement] ratification, leaders warned men to avoid new trouble” (5/7/38, 1).

Later that month, the Journal featured the AFL chief William Green denouncing the

CIO, ”In some cases industry has been plagued by sitdown strikes and the AFL and

62

its unions do not and will not condone such illegal methods of organization” (Ibid

5/22/38, 1).

The captive mines dispute was front-page news in the Milwaukee Journal five

times over the course of the episode. Here, the coverage was largely synoptic. The

most unfavorable lede was a quotation of a presidential statement wherein FDR as-

serted that ”the cessation of coal production would create a further danger to Amer-

ican defense and it was the indisputable obligation for the President to see the coal

production not be stopped” (Ibid 11/14/41, 1).

Overall, the Journal was not distinct from the national sample analyzed above

while the Deseret News contained examples of strike coverage that contextualized

the labor unrest. This is consistent with the observation that Utah was the only

state to keep its protective law intact. Sadly, these newspapers are not archived in

a fashion that allows for reliable keyword searching that would enable valid content

and volume analysis. However, in light of the relatively light coverage of strikes and

strikers by the Latter-Day Saints’ preferred newspaper, it is plausible to suggest that

the readers of Utah were situated in a distinct information context. Though including

more articles in the present analysis might be enriching, the core conclusion that the

national coverage was unfavorable would not be gainsaid. The New York papers are

treated in the main body of this chapter, but it is evident that the coverage there

was not favorable to strikes or strikers.

Manipulation or Education?

Even if the probes above were to persuade the reader that the press was pro-

capitalist (or anti-strike), it would not be clear whether this was simply to sell news-

papers or was revelatory of an editorial bias. A Herald Tribune editorial said the

63

Flint strike ”cannot be permitted to prevail if the law and the courts are to survive”

(1/10/37, 10). The Washington Post summarized its stance thusly, ”Sloan [GM chief]

is everlastingly right in demanding that his company’s plants be evacuated by un-

welcome squatters. There isn’t an honest, intelligent mind in the country that can

gainsay this. Least of all should it be condoned, even by silence, by the President

of the United States” (1/31/37, 9). The Christian Science Monitor was agnostic

by contrast, stating that ”The American conscience is still weighing the sit-down

strike” (2/1/37, 18). The Atlanta Constitution, though, adopted a law-and-order

stance, ”The grave threat of the sit-down strike to the welfare of the nation is causing

concern...the law is being flouted by strikers who refuse to respect authority in any

quarter” (3/23/37, 6).

In the captive mines dispute, the Washington Post submitted that ”John L.

Lewis has taken a stand in defiance of the United States. By his arbitrary refusal

to abandon coercion during this critical period, he has laid before the government

a challenge which it cannot fail to accept” (10/27/41, 10). And the Herald Tribune

took a similar tack, ”John L. Lewis’s peremptory defiance of the President in order to

call forthwith a strike which grabs the defense industry by the throat at least accom-

plishes this much - it puts the issue of the union shop...on the worst possible ground”

(10/27/41, 14). An editorial in the Atlanta Constitution argued that the strike ”has

forced a situation where the issue remaining is simply whether Lewis as head of his

union, or President Roosevelt, as head of the nation, shall dictate the terms on which

this nation’s defense and war aid program shall proceed” (11/18/41, 6). The New

York Times was less hyperbolic but was nonetheless worried about the power of the

United Mine Workers, pointing out that to award the workers a closed or union shop

in mediation would be ’capitulation’ (10/31/41, 22). The following week, they saw

that ”An increasing number of union leaders are no longer content merely to bring

64

pressure on the employer. They are attempting by the threat of strikes to coerce

government itself” (Ibid, 11/5/41, 22). The Los Angeles Times described Lewis as

a ’man up a tree’ and claimed that ”no ’labor leader’ since Debs has gone so far to

defy the government or flout the national interest as he” (11/14/41, 4).

In the Reconversion episode, the editorials were less strident, though they re-

mained opposed to strikes. From the Herald Tribune, ”The prospect that reconversion

to a peacetime economy will be hampered by labor strife is grave enough to demand

statesmanship...and the Wagner Act does not provide the machinery for a really fair

adjudication of the issues that will arise” (7/16/45, 8). The Washington Post was

dismayed to observe that ”no sooner had the news of Japan’s surrender been broad-

cast than a message from the CIO went over the airwaves stating that the no-strike

pledge was o↵” (8/18/45, 4). The Atlanta Constitution described the nascent strike

wave as ”spreading like an epidemic” (9/26/45, 6). Some editorials called for labor

peace, ”The President’s directive to the national conference of unions and manage-

ment, to find some way of settling disputes without strikes and without government

interference is good advice” (Los Angeles Times 11/6/45, 4). As the strike wave

wore on, the Washington Post editorialized, ”The immediate inconvenience caused

by any kind of large-scale labor disturbances is annoying rather than serious. But

the malign e↵ects of strikes and lockouts that slow down industrial activity and tend

to perpetuate wartime shortages of essential civilian products can hardly be exag-

gerated” (1/7/46, 6). Nor did the boards take the demands of unions seriously, the

Wall Street Journal was typical in this regard: ”President Philip Murray of the CIO

and past leaders of that organization are masters of drawing red herrings across the

path of plain logic and common sense” (1/30/46, 6). Overall, though, throughout the

year of 1946, the editorial emphasis was on the importance of production and less on

65

denouncing unions and strikes.

The editorial pages during the major strike episodes suggests that the editors

of the major dailies were decidedly unsympathetic to labor grievances, let alone labor

militance. Nor was this pattern confined to the post-Wagner period. ”No man may

be legally compelled to work for wages he does not consider adequate and no man

has the privilege of denying his neighbor the right to sell his labor at any price he

desires” (Washington Post 8/25/32, 6). The Christian Science Monitor embraced

a related theme, ”in every case of industrial warfare, the consumer is inevitably

the su↵erer...the cessation of any essential service cannot but strike at the innocent

bystander, the public, which stands all the inconvenience and pays all the bills”

(1/28/33, 16). Further, the Wall Street Journal approvingly referred to Senator

Wagner (D-NY) who ”told the American Federation of Labor in convention that

labor’s obligation was to maintain discipline within its own ranks and to win the

support of the public. In his mind public sympathy and support are not to be won

by the propensity so frequently of late to strike” (10/12/33, 8).

The editorials suggested that the patterns observed in the content analyses were

borne out of an anti-strike slant. However, this observation is not su�cient to label

the coverage of strikes as manipulative. The suggestion of slant does not demonstrate

the presence of manipulation. But it does rule out the possibility of labor militance

receiving a balanced hearing. Having said this, it is more likely that the outcomes

that interest this dissertation are likely as not the result of majority preference being

represented. Again, for the condition of manipulation to obtain, the editorial boards

would have had to have willfully intended to deceive. The observed patterns do not

rise to this level.

Conclusion

66

Taken together, these initial empirical probes combined with the next chap-

ter outline the contours of the possible in the mass polity. These alternate sites of

investigation have the advantage of revealing paths of potential variation that were

foreclosed. New Deal liberalism was delimited in the mass public at the height of

labor power - and militance - across the United States. Across jurisdictions, and

in advance of the conservative Democrat rebellion against the New Deal, the states

adopted laws to restrain strikes because ordinary citizens opposed the e↵orts taken

by other citizens that challenged existing relations. The relational threat posed by

militant industrial unionism demanded a powerful response.

The CIO saw itself as making a positive gain in terms of providing grievance

machinery. The CIO said that ’Patriotism outweighed all other considerations” (CIO

News June 1945). The CIO was flummoxed by a conundrum: the majority of people

are wage earners, so, then, how could the CIO be seen to seek special privilege?

Its interests paralleled the community’s since wage earners’ families largely make up

the community. Furthermore, the CIO took credit for the independent activities

of the Industrial Union Councils in New York and New Jersey that agitated for

the adoption of the first-in-the-nation fair employment legislation at the state level,

presaging its alliance to push the conservative Southern Democrats out of the caucus

(Baylor 2013).

In the present discussion, media owners are theorized to be motivated by the

desire for profit and support of free enterprise and not necessarily try to use their

power to influence the shape of public discussion and therefore use information as a

tool of class power and oppression. Political beliefs do not automatically reproduce

themselves nor does a social and economic order automatically reproduce itself. It

involves conscious propagation to sustain shared beliefs and values on a widespread

67

basis.

Consistent with Parenti’s observation (1986, 94-5) that protesters are described

more than their grievances are attended, the ways in which protesters were di↵erent

than the mainstream were widely emphasized in strike coverage. In part, for this

reason the landmark sit-down strikes were widely framed as an insurrection. By

framing the protests outside the mainstream, journalists served to reinforce the view

of the world that the capitalist ordering of the society and economy stood beyond

reproach.

The labor movement had great di�culty explaining its demands to the wider

public in a fashion that obscured their particularism. Consulting the wartime strike

record, one finds that workplace stoppages were the function of shopfloor conflicts

more than of organizational directive (Swo↵ord 1946, 9-14).10 Over the course of the

Wagner to Taft-Hartley period, the strikers were acting with an evidently di↵erent set

of motivations over time in their militance. At the outset of the New Deal, the strikes

were in furtherance of organization; as the New Deal wore on, the strikes came to be

motivated by shopfloor power struggles. In principle, this might frustrate the analyst

because the object of public derision is a shifting target within my temporal frame.

This is a challenge for inferring the underlying logics of mass opposition to strikes,

but it does not inhibit understanding the limits of progressive demand-making.

Strikes were essential to the gains the workers won through New Deal protest

- though this was not su�cient; the su�cient additional condition being the non-

intervention by the coercive apparatus (i.e. the National Guard not being used as

strike breakers). This is an important factor to consider for the labor movement was

subsequently incorporated into the Democratic coalition from a position of weakness.

10This was a study commissioned by Senator Joseph Ball (R-MN) of Ball-Burton-Hatch fame.

68

This was resultant from of the absence of other plausible coalition partners.

The coverage of strikes did seem to grow less favorable over the course of the

1930s, but it is far from evident that the news coverage was su�ciently unfavorable

to have constituted clear bias against labor. Unfavorable coverage does not imply un-

fairness, nor does it imply that public opinion was crafted by these articles. However,

the coverage certainly did not facilitate the formation of pro-strike opinion. And,

although the ’hard news’ regarding strike actions was factual, the choice was made

to focus more on the social costs of the strikes as opposed to the class conflict that is

inherent to the shop-floor. The simultaneous and nearly instantaneous retrenchment

against strikes suggests that these phenomena were governed by a shared logic. The

next chapter turns to survey data that depict the attitudes citizens held in the light

of the coverage analyzed in this chapter.

69

Chapter 3

What Did the Public Think of

Strikes?

Introduction

In the previous chapter, I investigated the coverage of strikes in newspapers

across the nation and in the present chapter, I turn to surveys of the citizens that

elected anti-strike legislative majorities across the nation. What did citizens think

about strikes and would it meaningful or sound to understand them as a homogeneous

public? If citizens’ views on militance were sharply divided along class or regional

lines, then the adoption of Taft-Hartley might be an instance of class or regional dom-

ination. On the other hand, if citizens were responding to the emerging turbulence in

a similar fashion, this would be consistent with the view that the public got what it

wanted from New Deal labor law. Finally, to what extent was mass opinion in tension

with the coverage of strikes presented in the previous chapter?

It is puzzling that the public came to take a dim view of industrial organizing,

70

rather than blaming capital. The public disfavor of the CIO strikes appears wrong-

headed in that most citizens had class interests more in common with the strikers

than with the owners. Interests are a treacherous topic from a normative or posi-

tive perspective. Defining an interest necessarily implies a discounted utility function

whose discount rate is impossible to demonstrate. If the rate is posited a priori, then

one has a di�cult time demonstrating that observed preferences that di↵er are flatly

wrong. Nevertheless, it is equally di�cult to posit that a political vehicle in American

history has achieved more gains for economic losers than organized labor. And the

power of labor over capital lies not in the fact of its organization, but in the withhold-

ing of its productive capacity. Moreover, the very process of profit-generation implies

exploitation: in order for a firm to make a profit, it must sell its goods for more than

it costs to produce them; a positive di↵erence can only arise when labor earns less

than it creates.1

The negative potency of union militance was also reflected by the media cov-

erage of strikes for ”the symbols of revolt are not drawn like musty costumes from a

cultural closet and arrayed before the public. Nor are meanings unrolled out of whole

cloth. The costumes of revolt are woven from a blend of inherited and invented fibers

into collective action frames” (Tarrow 1998, quoted in Kellstedt 2003, 83). The 1930s

strikes were evidently highly salient and, though citizens are typically disinterested in

news, they can typically be counted upon to perk up when it clearly matters to them

(Key 1965; Page and Shapiro 1994). As noted in the introduction, this dissertation

di↵ers from much of the New Deal literature not only because of its focus on the

1To point out that the capitalist had secured unpaid labor for a sold product does not mean thesame thing as to say that he has secured a profit. Elaborating beyond the above simplification wouldtake us beyond the scope of the present study. In future work, though, I would like to explore therelationship between marginal profit, workplace tyranny, and political activity; it is my suspicionthat firms with narrower margins tend to be more extreme in their political rightism and tend tovisit more depredations on their workers.

71

states, but also its focus on bottom-up policy change. Here we are interested in how

much the substance of citizens views were connected with newspaper content.

To buttress the rationale for studying newspapers, I examined Gallup poll re-

sponses to the question ”Do you read a daily newspaper?” In April 1937, across

regions and class, more than eighty percent of respondents a�rmed that they read

a newspaper on a daily basis. In October of the same year, the same result held.

And, once again, in April of 1939, the overwhelming response was in the a�rma-

tive across classes and regions. Though it is worth pointing out, that as one might

expect, the South and Sunbelt were the least-read regions by self-report. Desirabil-

ity bias raises measurement validity concerns but this result dovetails with Bartels’

(1993, 269) observation that newspaper reading ”appears to be a su�ciently stable

behavior.” Furthermore, assume the contradictory position: if citizens formed views

on strikes from personal experience, then there should have been clear regional varia-

tion in opinion as the strikes were geographically concentrated. In May 1939, Gallup

asked ”Would you prefer to get national news from a daily newspaper or from the

radio?” Interestingly, there was a class and regional cleavage here wherein the upper-

class respondents preferred newspapers; and, the Midwest (West Central) and the

Mountain West expressed an equal preference for print and radio. Sadly, this poll did

not contain a question asking whether the respondent actually owned a radio because

the variation in class preference might be a result of deprivation.

Happily, in 1944, we can explore responses to the same question juxtaposed

with radio ownership. By 1944, respondents owned radios by a greater than 4:1 ratio

and at this point across class and regions, citizens preferred to get their news from

the radio broadcasts. The turn toward radio and then television coverage carries im-

plications for exposure and processing that are distinct unto themselves (Prior 2006,

72

for example), but that is beyond the scope of this investigation. Some studies have

suggested that radio content significantly overlapped with newspaper coverage (Albig

1937; Anderson 1971, for example) and that the radio coverage was mostly unfavor-

able in its coverage of labor unions (Sussman 1945; Schiller 1981, for example). As

we will see, the backlash to strikes and strikers did antedate the growing preeminence

of radio.

In outlining the contours of citizen attitudes toward unions, strikes, and the

Wagner Act, I find that, contrary to expectations drawn from sectional or class in-

terest, the mass public was generally hostile to worker protest upon the formation of

the New Deal coalition. In principle, the lower-class citizens and the more urbanized

regions might have had more sympathy for unionists and strikers.

When citizens consumed news about strikes, what kind of news did they con-

sume? This is an important question because the strikers were struggling to be

understood by the mass polity. The CIO actively sought to shape interpretations

of events though speeches, radio programs, and pamphlets (Foster 1975; Fones-Wolf

1994; Charnock 2012). In surmising why citizens might have held similar opinions, we

would look at the information they shared. Moreover, since Hadley Cantril, observers

and students of politics have known citizens to make their evaluations of policy more

on a concept of the public good than their own self-interest; and to rely on elite cues

to form standards and evaluations (Eisinger 2003). Elite cues comprise much of the

substance of the news and citizens were typically exposed to a limited number of

sources. So, then, subgroup opinion, having been exposed to information in common

and sharing processing rules, would tend to move in parallel (Page and Shapiro 1992,

318-9).

In this chapter, I assemble aggregated national survey responses to items re-

73

garding labor militance, particularly the sit-down and strikes during the run-up to

the Second World War. Previous studies identifying a mass backlash to the CIO have

emphasized the importance of the post-War strike wave in giving the Taft-Hartley

delegation a workable majority in both chambers of Congress (Lee, 1947 47-52; Atle-

son 1998, Chap 11, for example). In this chapter, I will identify mass opposition to

union power that antecedes the War, let alone the Truman presidency. Patterson

(1969) identified the rise of the conservative majority in the Congress as a result of

backlash to court-packing in part. I will also examine public attitudes on this.

Classic works in political science emphasize that the impact of interest groups

is directly related to the esteem in which they are generally held (Key 1964; Truman

1971). Therefore, it is necessary to outline the contours of public opinion on unions

and strikes in order to see whether the glass is ’half-empty.’ There was evidently

more tolerance for the disruptive strikes for union organization than for the strikes

to secure particularistic benefits, political clout, or programmatic gains.

Typically, political science studies of the New Deal are focused on the politi-

cians and the bureaucracy, but the mass public was a target of e↵orts at persuasion.

Political elites - particularly the president and newspaper owners - were interested in

shaping opinion within the hard bounds set by mass preferences. This task is essential

to delimiting the limits of liberalism: because the white supremacist delegation had

a veto tells us why Congress was not a potential site for improving labor’s statutory

fortunes; that senior liberals in the FDR administration were not committed to code-

termination of production tells us why the bureaucracy was not a potential site for

improving labor’s regulatory fortunes. But these obstacles were themselves held in

place by a bulwark of mass opposition to labor defiance. More subtly, this chapter

sets the stage for connecting individual-level attitudes with the legislative outcomes

74

analyzed in the next chapter that interest this dissertation.

The public opinion literature that utilizes relevant polling from the period

(Verba and Schlozman 1977; Berinsky and Schickler 2006; Panagopoulos and Francia

2008; Schickler and Caughey 2011) has broadly covered this ground before, but the

aggregate stability of opinion is left unconnected to the development of the labor

relations framework. Implicit in the backlash concept is the notion that the mass

polity held unions in disfavor. This empirical question will be explored as it has been

in previous e↵orts (Bok and Dunlop 1970; Parenti 1986; Schmidt 1993; Flynn 2000;

Martin 2004; Carreiro 2005; Amenta et al. 2009, for example). Thus, this chapter is

not only interested in using this analysis to ask whether the organized labor move-

ment might have achieved a better result from the New Deal; it is interested in the

contours prevailing attitudes toward militance and the protective laws that promoted

it. In particular, this chapter is distinctive in its focus on salient strike episodes and

its emphasis on pre- and post-War continuity in mass opinion.

In studying the volume and content of news, I have been implicitly considering

opinion formation at the individual level; however, the survey data examined here are

treated as aggregates. The data are not ideal but they are better than what students

had even a decade ago and they are essential for bringing in public opinion. And

the question is not whether the data are ideal but whether we can plausibly proceed

without it. If public support for strike restrictions is not noted, then we would not

see why the Taft-Hartley restrictions were ineluctable.

Again, the exclusionary focus on labor militance as opposed to alternate eco-

nomic or social items is motivated in part by the observation that capital was satisfied

with seeing unions denuded over the course of decades in the cul-de-sac of federal la-

bor law since Taft-Hartley (Estlund 2002). More to the point, the labor relations

75

area is one of the few areas in which mass opinion moved rightward during FDR’s

tenure (Schickler and Caughey 2011, 163-4). And the stability and breadth of re-

strictive opinion helped to ensure that the labor-liberal alliance would be checked

until it could make an alliance with the civil rights demanders to take control of the

Democratic Party (Frymer 2008; Roof 2011; Baylor 2013).

In addition to being crucial in the development of the American labor relations

order, the strike waves of the New Deal period lends an opportunity to examine

conflict between broad economic groups in society. Particular attention to the sit

down wave of 1937 and the UMW (Lewis’ group and the financial core of the CIO)

disputes of 1941 will reveal that the contours of mass opinion did not hold any promise

for a more favorable legal outcome in the United States.

Theory and Motivation

The foregoing analysis has taken 1937 as the starting point for the investigation

because Wagner was ratified as legitimate in that year, which allowed the states to

play out the labor relations development alongside the federal government. Yet,

during the same year, the anti-labor coalition of Northern Republicans and Southern

Democrats began to emerge as President Roosevelt sought to pack the Supreme Court

and ’primary’ conservative Democrats (Patterson 1969). Amidst this fracturing of the

Wagner coalition, the anti-labor laws that would structure the labor relations order2

for the rest of the century into the next were adopted. The chambers that enacted

the restrictive backlash came to power in the 1938 elections, thus an account of the

backlash would be incomplete without reference to public opinion.

Though the search for class cleavages in mass opinion toward strikes is the chief

2As before, ’labor relations’ connote the shop-floor dealings of employer and employee. ’Order’means overarching framework or legal structure.

76

motivation for this chapter, regional cleavages are of interest as well. Following Bensel

(1984), I distinguish between the industrial core and periphery. In this seminal work,

Bensel argues that sectional interests were decisive over bill adoption across a variety

of economic issues, as opposed to partisanship or ideology. In the present chapter,

we might expect to see that worker protest was more acceptable to respondents in

the industrial core than to those in the periphery because core respondents might

have been more likely to identify their common cause with the strikers. Alternately,

presence in the industrial core likely exposed one to the inconvenience related to labor

stoppages, thus increasing the likelihood of opposing worker protest. At all events,

the argument that sectionalism has been the dominant factor shaping the political

economy implies substantial regional variation in opinion unless strikes were widely

understood as an injury to all. For the purposes of the regional classifications below,

we will take the South and Southwest and Mountain West regions as the periphery

while the other regions will be taken to be in the core. Though this approach does

not have the granularity of Bensel’s ’trade areas’ taxonomy, it will be su�cient to

examine whether opinion was determined by the respondent’s region.

This chapter will show that there was support for strike bans as well as a

generalized opposition to militant unionism across the nation amidst muted class

cleavages. As we will see in the next chapter, the immediate upshot was Republican

seat gains in 1938, particularly in the states that adopted ’baby’ Wagner Acts. In

short, the industrial union organizing precipitated by Wagner caused a social and

political perturbation that was programatically threatening but insu�ciently strong

to repel the backlash.

The term ’backlash’ implies that there existed a previous equilibrium that had

been disturbed (Pierson 2000, 85). Before the Wagner Act, would-be union militants

77

had little reason to expect that statutory gains would be e↵ective in the face of judicial

opposition. The Supreme Court upholding Wagner in the face of political pressure

legitimized strike activity. Consequently, the newly legitimized industrial unions were

wedded to the Democratic Party, because it was the Democratic administrative com-

mitment to Labor Relations Boards that made the CIO growth legal. But the direct

e↵ect of CIO growth positioned the industrial unions in direct opposition to prior

members of the coalitions that enacted, or might have enacted, protective laws. And,

Gross (1981, Chapter 2) rightly emphasizes the role that public criticism of the NLRB

had in the push for Wagner emendation. The Supreme Court held, consistent with

prevailing views, that the illegal seizure of plants by strikers was ”more troublesome

than the illegal actions of employers” (Ibid, 26). And, further, though Roosevelt was

popularly connected to the Wagner Act, the president himself was skeptical of the

Board’s pro-CIO leaning (Ibid, 39-40).

There was a shift in the legitimate exercise of authority in labor relations (cf.

Orren and Skowronek 2000) as a result of Wagner. Prior to this, the dominant

pattern of labor relations was a relation of force. This substantial change in policy

seemed to contemporary observers to indicate that a fundamental social reordering

was nigh (Goldfield 1989, 1264-5; Columbia Oral History Project, Frances Perkins

Reminiscences). With the benefit of hindsight, it is easy to understate the peril that

was perceived by elites. This perceived danger forms the first empirical question of

this chapter: how class conscious might the public have been? Put another way, did

the poorer respondents identify with union resistance?

Plausibly, one might suggest that personal experience with strikes would be

su�cient to form a negative opinion. For example, one’s having been deprived of

milk due to a transport sympathy strike with the dairy cooperative could well have

78

turned one’s opinion against CIO organizers. However, as we will see, the strikes

were regionally concentrated and mass opinion does not show meaningful regional

variation. Moreover, though the responses show clear di↵erences in levels of opinion

by occupational class suggesting that opinion formation was a function of interests, the

parallel movement of class opinion over time is consistent with Verba and Schlozman’s

(1977, 317) conclusion that New Deal respondents lacked class-consciousness broadly -

indeed, majorities of blue-collar and unemployed respondents did not see management

and labor as being in opposition (Ibid, 308). Like that study, I find little evidence that

the unemployed or the poor were particularly desirous of robust industrial democracy,

shop floor codetermination, let alone ending capitalism.

’Switch in Time’

The court packing episode was a turning point in the history of the Wagner Act

and is deeply relevant to the present analysis. The Supreme Court had held several

New Deal measures unconstitutional, including the collective bargaining provision of

the Norris-LaGuardia Act which the Wagner Act was designed to replace. The Court

was expected to rule the Wagner Act unconstitutional as well. This episode is deeply

relevant because the President might have been seen as to have been subverting old

norms in order to reward his new friends in labor. The backlash under study here

may have been against the perceived presidential overreach rather than the industrial

union disruption.

Happily, these survey data present the opportunity to revisit the 1937 court

packing controversy. Early in 1937, FDR threatened to pack the Court in the face

of its doctrinaire constitutional reading that denied the right to withhold labor while

enjoining capital to bargain. It is important to note that interpreting public opinion

on the court packing controversy and labor issues is fraught with pitfalls as there

79

was a long chain of reasoning linking the Court and the right to strike. And, the

court-packing plan was about more than the Wagner Act; it was about the power

relations between the branches.3

Though the Supreme Court had, at various points in US history, been the

target of fulminations by sitting presidents and public outrage, the court packing

moment stands out as consequential because it coincides with the beginning of the

end of New Deal liberalism. However, the proximate cause for the proposal was the

anticipated rejection of the Wagner Act after the Court had found previous elements

of the New Deal unconstitutional, particularly matters related to labor relations. It is

plausible that the key to union legislative defeat was that the Supreme Court ratified

the Wagner Act under duress. That is, FDR subsequently betrayed the militant core

of the CIO that was central to his reelection because he had spent all his ’political

capital’ protecting the Wagner Act (which he did not support openly until a few weeks

before its passage).4 Or, one might argue that mass disfavor of the union power by

which the Supreme Court was cowed also generated the electoral backlash to FDR

in the Congress and in the states that made leftward movement in New Deal policy

nigh-impossible. At all events, the Roosevelt reelection coalition included numerous

Republican supporters (Sinclair 1978; Erikson and Tedin 1981), which suggests that

the mass support for further leftward policy change was limited, particularly a shift

at the cost of long-standing US constitutional patterns.

3Perhaps it is worth considering, despite being speculative, that FDR might have been able toenlarge the Court had he not been insistent on adding six judges. According to Altman, the Senatewas going to recess daily and limit speeches in opposition to the bill (1937, 1079). That is to saythat Roosevelt might have successfully attained Senate approval for enlarging the Court had he notsought to emerge with such an overwhelming majority of co-partisans.

4The ’political capital’ metaphor, though often tenuous, is particularly inapt in the presentinstance. What was happening during this 1933-1937 period punctuated by general strikes in popu-lation centers and sit-down strikes across core industries constituted a direct assault on the existingsocial relations, and, perhaps, the right to property altogether. (Lichtenstein 1997)

80

Caldeira in his study of Gallup polls from 1938 (1987, 1149) finds that ”two

crucial events - Jones and Laughlin and Justice Van Devanter’s resignation - spelled

doom for FDR’s bill to enlarge the Court and pack it with justices favorable to the

New Deal.” That is, the acquiescence of the Court to the demand that the Wagner

Act be upheld and the stepping-down of one of the judicial opponents of the New Deal

cooled public enthusiasm for altering the Court’s structure. However, this influential

work did not present subgroup analyses, preferring to treat the issue items as a

continuous dependent variable in a linear model. Here, I am more interested in the

class and regional cleavages that theory would expect: the upper class, being aware of

the Court and its conservative role ought to have disfavored enlargement; and those

on relief who owed their presence on the rolls to Roosevelt would have tended to

support the president. Baum and Kernell (2001) find evidence that cleavages in class

approval of the Roosevelt presidency were a peacetime phenomenon that was eroded

by US entry into the Second World War.

FDR, having been a highly salient figure, obscures the interpretation of citizens

views because approval of the politician himself conditions responses on items related

to him. But the instance was a turning point in his presidency as FDR pivoted toward

foreign policy thereafter. Of course, the loss of liberal Democratic seats in the 1938

elections played a significant role of Roosevelt’s calculus. Although the Sunday New

York Times of April 3, 1938 pointed out that ”The sea and things of the sea, the

navy and its ships and men and guns are probably the outstanding passions of the

President’s life” and ”The Navy is being run from the White House these days.” It is

hard to say whether the patrician former Undersecretary of War for the Navy would

have sought to pivot even if he had packed the Court with cronies and had endless

coattails (or, at least, not reverse ones). Moreover, the attempt to ’purge’ recalcitrant

81

Bourbon Democrats did actually succeed in unseating John O’Connor, the chairman

of the Rules committee in the House, and an opponent of liberal New Deal measures

(Polenberg 1968).5

Data and Method

In the following, I ask whether class and region put di↵erent groups at odds over

the legitimacy of strike actions. Though limited by available survey data, I present

subgroup opinion (see Appendix for question wordings) on salient items during 1937,

the height of the sit-down troubles. I continue to dwell on the important year of 1937,

wherein Roosevelt proposed to enlarge the Supreme Court. Then, I present opinion

responses to the union approval item form 1937 to 1943 comparing respondents by

region and by class. I turn to opinions on the Wagner Act, which was a salient issue

throughout the period; then opinions on John L. Lewis the militant labor leader; and

opinions on defense strikes during the run-up to the Second World War.

In order to assess public opinion regarding militance in 1937, I take a collection

of issue items from 1937 Gallup polls and present the subgroups comparisons by re-

gion6and class7 However, there are no items repeated across all the 1937 surveys, save

FDR approval questions. Moreover, to the extent that the items are heterogeneous,

5Though the result was the ascension of Dixiecrat Howard Smith of Virginia.

6The regions are as follows: Northeast (Maine, New Hampshire, Vermont, Massachusetts,Rhode Island, Connecticut), Middle Atlantic (New York, New Jersey, Pennsylvania, Maryland,Delaware, West Virginia), East Central (Ohio, Michigan, Indiana, Illinois), West Central (Wis-consin, Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, Kansas), South andSouthwest (North Carolina, South Carolina, Virginia, Georgia, Alabama, Arkansas, Florida, Ken-tucky, Louisiana, Mississippi, Oklahoma, Tennessee, Texas), Mountain West (Montana, Arizona,Colorado, Idaho, Wyoming, Utah, Nevada, New Mexico), and Pacific Coast (California, Oregon,Washington).

7The classes were Average-plus (Upper), Average (Middle), Poor and Poor-plus (Lower), andthose receiving emergency aid (Relief). The Gallup interviewer scored the respondent as belongingto the economic classes having screened for those reporting having received relief spending. Thoughthis is seemingly a recipe for coder bias problems, the recorded scale correlated with reported incomeat .74 (Baum and Kernell 2001, 205).

82

this serves to bolster the claim that regionalism was subsumed into a broader opinion

climate that was itself mediated.

Despite their distinctiveness, the items address the substance of the CIO threat

directly. The personality of the leader John L. Lewis was salient because he was

famous for his militance and the Mine Workers provided the bulk of the CIO’s op-

erating budget (Foster 1975) and the sit-down questions bear on the current study

directly. The key point here is that the class cleavages in the mass public, then as

now, were not su�cient to inspire counter publics. Workers qua citizens did not favor

radical leftist programs, and proved receptive to messages that constructed unions as

negative forces.

Though it may seem curious not to dwell on the union approval series, this

ground is covered well by Schickler and collaborators in previous e↵orts. What these

studies do not emphasize - and what we touched upon in the previous chapter - is

that the dips in union approval from their all-time highs in 1940 and 1954 (Gallup

2011) were precipitated by racketeering scandals in AFL unions. And indeed, even

Robert Taft might have answered the union approval question in the a�rmative at

various points throughout the New Deal; what was critical was the widespread view

that union militance was excessive.

Analysis

The union approval opinion response did not vary systematically by region,

as shown in Figures 3.1 to 3.3. The regions moved roughly in tandem over time. In

Figure 3.1, the Northeast, Mid-Atlantic, and Pacific Coast regions are plotted against

the overall response. I put these regions together because these states were all in the

83

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Aug

-37

Jan-

38

Jun-

38

Nov

-38

Apr

-39

Sep

-39

Feb-

40

Jul-4

0 D

ec-4

0 M

ay-4

1 O

ct-4

1 M

ar-4

2 A

ug-4

2 Ja

n-43

Per

cent

App

rove

Date

Northeast

Mid Atlantic

Pacific Coast

Overall

Figure 3.1: Union Approval, Northeast, Mid-Atlantic, Pacific

industrial core. The Northeast moved in tandem with the rest of the nation, though

these states did typically approve of unions at a higher rate than average. Even at

the troughs of support, a majority of the public expressed a favorable view of unions;

and notably, these troughs are found in the strike-heavy years of 1937 and 1941.

No less than 60 percent of the public expressed a favorable view of unions in this

region, or the nation generally. The Mid-Atlantic region, which has the highest level

of union support, consistent with those states being in the heart of the industrial core;

however, the region does move in tandem with the other regions. Absolute levels of

union support were typically greater than 70 percent and not below 65 percent in

any survey. Perhaps unsurprisingly, the Pacific Coast (Figure 1), with strong left-

wing contingents in their shipyard unions, contained the most favorable response to

unions, approving at greater than the national average in each survey, though moving

in parallel.

84

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Aug

-37

Jan-

38

Jun-

38

Nov

-38

Apr

-39

Sep

-39

Feb-

40

Jul-4

0 D

ec-4

0 M

ay-4

1 O

ct-4

1 M

ar-4

2 A

ug-4

2 Ja

n-43

Per

cent

App

rove

Date

East Central

West Central

Overall

Figure 3.2: Union Approval, East Central, West Central

The East Central region, as depicted in Figure 3.2 closely tracked the national

average response. However, the West Central region consistently approved of unions

at a below-average rate. I juxtapose these regional trends because these states were

in the industrial core, but were marked by higher levels of agricultural production.

Also in this figure, we observe parallelism in the direction of opinion change in each

survey but this region tends to disapprove of unions more than average. Both regions

maintain 65 percent approval at the lowest survey readings, though they average 74

and 72 percent respectively.

Figure 3.3 depicts the response for the South and Southwest alongside the

Mountain West; unsurprisingly, these periphery regions had the only net negative

response to the union approval question during the captive mines dispute in 1941

wherein the South reported a 47 percent approval rating of unions. On average,

however, these regions average 64 and 67 percent respectively. Although the states in

85

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Aug

-37

Jan-

38

Jun-

38

Nov

-38

Apr

-39

Sep

-39

Feb-

40

Jul-4

0 D

ec-4

0 M

ay-4

1 O

ct-4

1 M

ar-4

2 A

ug-4

2 Ja

n-43

Per

cent

App

rove

Date

SSW

Mountain West

Overall

Figure 3.3: Union Approval, South and Southwest, Mountain West

this region were distinctive, not only in terms of Jim Crow laws but also in terms of

military siting, the responses change in a parallel fashion. The Mountain West whose

response closely tracked the national average, with levels of approval not going lower

than 65 percent.

Consistent with the present day, the South and Southwest region consistently

reported the lowest relative approval of labor unions while the Northeast was generally

the most favorably disposed. Notably, while the South declined in union approval

during the captive mines dispute, the Mountain West did not. Within each item, there

is consistent homogeneity across regions. Also, none of the regions are consistently

outlier subsamples. The patterning suggests that regionalism did not have a strong

e↵ect on opinion related to union approval. In terms of temporal di↵erences, there

was a generalized dip in union approval in 1938, preceding the electoral backlash to

FDR.

86

Each region is positively correlated with every other region on union approval.

The Mountain West was the most distinctive region with respect to union approval

compared to the other regions. The Mountain West was most weakly correlated with

the Northeast and the Mid-Atlantic, consistent with the argument that states in the

core and periphery would express di↵erent attitudes. The public in Utah displayed

weaker support for unions than in the Northeast, yet the Utah law remained intact.

However, the overall level of union approval in this region closely tracked the national

average. And, moreover, unions enjoyed a net positive approval rating in each survey

outside the South. Union approval levels were typically highest on the coasts dotted

by immigrant-rich cities.

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Aug

-37

Jan-

38

Jun-

38

Nov

-38

Apr

-39

Sep

-39

Feb-

40

Jul-4

0 D

ec-4

0 M

ay-4

1 O

ct-4

1 M

ar-4

2 A

ug-4

2 Ja

n-43

Per

cent

App

rove

Date

Upper

Middle

Overall

Figure 3.4: Union Approval, Upper and Middle Class

Figures 3.4 and 3.5 depict the class sub-samples for the same union approval

items. In Figure 4, the responses from upper-class respondents are juxtaposed with

the middle-class respondents alongside the national average. These respondents, pre-

87

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Aug

-37

Jan-

38

Jun-

38

Nov

-38

Apr

-39

Sep

-39

Feb-

40

Jul-4

0 D

ec-4

0 M

ay-4

1 O

ct-4

1 M

ar-4

2 A

ug-4

2 Ja

n-43

Per

cent

App

rove

Date

Lower

Relief

Overall

Figure 3.5: Union Approval, Lower Class and Unemployed

dictably, were more likely to disfavor unions. Nonetheless, upper-class respondents

did express a net favorable opinion of unions across surveys with no survey reading

showing a lower than 55 percent approval. Here, though, the expectation of opinion

parallelism breaks down somewhat as the subsample opinion moves in opposite di-

rections to the national average, increasing during an overall decrease and vice-versa.

Though the middle-class respondents typically disapproved of labor unions more than

the national average, the change in survey responses moved in parallel with the na-

tional average. As before the troughs are in 1937 and 1941, but the readings are all

above 65 percent. These results do not suggest a class cleavage in union approval

response.

Likewise in Figure 3.5, the lower-class respondents moved in parallel with the

national average though approving of unions at level slightly higher than the national

average. Here the approval levels were not less than 65 percent. In this figure we see

88

that the respondents on the relief rolls approved of unions more than the national

average in most surveys and the subgroup opinion mostly moved in parallel with

the national average (r=.88). And, importantly, the poorest respondents expressed

a lower level of approval than the upper class respondents (r=.27) in 1937 and 1941.

While the unemployed respondents - the ones on relief - typically had the highest

approval of unions and the respondents coded as upper class held unions in the lowest

relative regard, this condition did not hold across all surveys, thus militating against

there having been a class cleavage in opinion. However, the general trend over the

1937 to 1944 period was downward. And, again, the distinct publics move in parallel

rather than in a fashion that is suggestive of class cleavage. Though citizens typically

lost regard for unions, they do not tend to view them in an unfavorable fashion. It

should also be noted that the union approval time series is at its lowest point prior to

2009 in 1941 (Gallup 2014). The upper-class respondents had the weakest correlation

with the average response among the class groups, while the other groups were close

together in approval levels. Importantly, though, the correlation was weakest between

the upper and lower-class, as opposed to with those on the welfare rolls. But even

the upper-class respondents did not report net disfavor in any survey. The key story

with union approval here is that though the overall levels of approval are high, the

trend throughout the critical early post-Wagner period was downward.

Figures 3.6 to 3.8 show the responses to the John L. Lewis survey items by

regional subsample. In Figure 3.6, the Northeast (r=.85), Mid-Atlantic (r=.66), and

Pacific Coast (r=.71) regions all closely tracked the national average and are generally

declining over time. Though, consistent with being in CIO strongholds, these regions

were consistently above the national average. And the Pacific Coast region, is likewise

parallel, though Lewis was unsurprisingly more popular in this region which was a

89

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Jan-

37

Jul-3

7 Ja

n-38

Ju

l-38

Jan-

39

Jul-3

9 Ja

n-40

Ju

l-40

Jan-

41

Jul-4

1 Ja

n-42

Ju

l-42

Jan-

43

Per

cent

App

rove

Date

Northeast

Mid Atlantic

Pacific Coast

Overall

Figure 3.6: John L. Lewis Approval, Northeast, Mid-Atlantic, Pacific

site of radical worker protest. Figure 3.7 depicts the East Central and West Central

regions, which also shows graphical evidence of parallelism.

Figure 3.8, depicting the South and Southwest(r=.49) and the Mountain West

also evidences parallelism, though this region expresses the lowest levels of Lewis

approval, as both regions expressed below-average Lewis approval. The average levels

of approval move from net favorable to deeply net unfavorable over the six years of

survey items. The national average was net 2 favorable in early 1937 and fell to

negative 80 percent after the captive mines dispute. Indeed, every region at the nadir

of Lewis approval during the captive mines dispute was below 20 percent.

Figures 3.9 through 3.10 depict the Lewis approval responses by class. Pre-

dictably, in Figure 3.9 we observe that the upper-class respondents were markedly

less likely to approve of John L. Lewis though this sub-sample was moving in tandem

90

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Jan-

37

Jul-3

7 Ja

n-38

Ju

l-38

Jan-

39

Jul-3

9 Ja

n-40

Ju

l-40

Jan-

41

Jul-4

1 Ja

n-42

Ju

l-42

Jan-

43

Per

cent

App

rove

Date

East Central

West Central

Overall

Figure 3.7: John L. Lewis Approval, East Central, West Central

with the national average (r=-.43). Also in Figure 3.9, we see that the responses

for the middle-class, whose opinion was close to the national average within and

across surveys. In Figure 10, the lower-class respondents, constituted a parallel pub-

lic as well, though, predictably, they supported Lewis more than the national average.

Likewise, the responses by those on relief, who typically supported Lewis above the

national average. However, importantly, this group which had the most to gain from

Lewis’ opposition to the political relations between the classes, converged toward the

generally high disapproval of Lewis by the run-up to the Second World War in light

of the ’captive mines’ dispute, which we treated in the previous chapter.

In general, the items specific to John L. Lewis showed a clear overall trend

toward increasing disapproval of the militant labor leader. Across regions and classes

the trend was identical. This is in keeping with the narrative of his increasing es-

trangement from the CIO and from the Democratic Party. Lewis had been regarded

91

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Jan-

37

Jul-3

7 Ja

n-38

Ju

l-38

Jan-

39

Jul-3

9 Ja

n-40

Ju

l-40

Jan-

41

Jul-4

1 Ja

n-42

Ju

l-42

Jan-

43

Per

cent

App

rove

Date

SSW

Mountain West

Overall

Figure 3.8: John L. Lewis Approval, South and Southwest, Mountain West

as a marginal figure as a result of his break with Hillman and Murray (Lichtenstein

1982; Zieger 1995); but the evident threat his vision posed to the incumbent Demo-

cratic Party suggests that views on Lewis implied views on the legitimacy of the ruling

coalition.8 Importantly, John L. Lewis approval began to plummet in 1937 leading

into the turning-point elections of 1938.

Regional averages in support of Lewis were mostly positively correlated, though

significantly less so than in the case of overall union approval. Notably, the West

Central region, containing Wisconsin, had negative correlation with the Northeast,

Mid-Atlantic, and Pacific regions. This highly agricultural region was opposed to

Lewis during the captive mines dispute and his subsequent move to organize dairy

workers.

8Also, we should note that Lewis’ notoriety ought not be conflated with leftism; indeed, he wassupportive of capitalism, even of the laissez faire variety.

92

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Jan-

37

Jun-

37

Nov

-37

Apr

-38

Sep

-38

Feb-

39

Jul-3

9 D

ec-3

9 M

ay-4

0 O

ct-4

0 M

ar-4

1 A

ug-4

1 Ja

n-42

Ju

n-42

N

ov-4

2 A

pr-4

3

Per

cent

App

rove

Date

Upper

Middle

Overall

Figure 3.9: John L. Lewis Approval, Upper and Middle Class

However, there was substantial divergence in class opinion on Lewis. The upper-

and middle-class respondents substantially diverged from the lower-class and unem-

ployed. Upper-class responses were not at all correlated with the other groups, sug-

gesting that the disruptive threat posed by Lewis was clearly recognized by those

who had the most to lose. Though every class displayed a negative net favorable

rating of John L. Lewis, the poorer respondents had a higher peak from which to

fall. But these di↵erences are fairly muted: upper-class respondents expressed a net

negative 94 percent approval, while respondents on relief expressed a net negative 78

percent.

In 1941, the identical question was put 6 times over the course of the year:

”Should the Government forbid strikes in industries manufacturing materials for our

national defense program, or should the workers in those industries continue to have

the right to strike?” Figure 3.11 depicts the level of support for measures to outlaw

93

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Jan-

37

Jun-

37

Nov

-37

Apr

-38

Sep

-38

Feb-

39

Jul-3

9 D

ec-3

9 M

ay-4

0 O

ct-4

0 M

ar-4

1 A

ug-4

1 Ja

n-42

Ju

n-42

N

ov-4

2 A

pr-4

3

Per

cent

App

rove

Date

Lower

Relief

Overall

Figure 3.10: John L. Lewis Approval, Lower Class and Unemployed

strikes in the Northeast, Mid-Atlantic, and Pacific Coast regions plotted alongside the

national average. The Northeast region moved in tandem with the national average

but this industrial core region supported such a measure at a markedly lower rate.

The Mid-Atlantic subset tells a similar story. Likewise, the Pacific states was more

muted in their support for a wartime strike ban.

The East Central region (Figure 3.12) which closely tracked the national aver-

age. The West Central trend shows that this region tended to support the strike ban

above the national average and this region supported it at an increased level when

the national average declined.

Figure 3.13 depicts the South and Southwest where support for strike bans was

consistently elevated above the national average, though this region moved in parallel

with the nation. The Mountain West which tended to support the strike ban above

94

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr-41

May-41

Jun-4

1

Jul-4

1

Aug-41

Sep-41

Oct-41

Nov-41

Dec-41

Per

cent

App

rove

Date

Northeast

Mid Atlantic

Pacific Coast

Overall

Figure 3.11: Strike Ban Approval, Northeast, Mid-Atlantic, Pacific

the national average though less so than the South and Southwest. No less than 65

percent of the public in any region expressed support for an outright ban on strikes

in defense industries. It is, furthermore important to note that any large industry

could have been easily construed as defense-related.

Figure 3.14 shows the upper-class respondents’ support for defense strike bans

which moved in tandem with the national average, though, unsurprisingly was above

it. The responses for the middle-class tracked the national average closely. Unsurpris-

ingly, the lower-class were more likely to oppose a ban on defense strikes, though a

decisive majority of those respondents favored one as shown in Figure 3.15, which also

reveals that the respondents on the relief rolls were the least likely to support such a

measure, though this group opinion moved in tandem with the national average

Unlike the above items, the regions are negatively related in several cases and

95

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr-41

May-41

Jun-4

1

Jul-4

1

Aug-41

Sep-41

Oct-41

Nov-41

Dec-41

Per

cent

App

rove

Date

East Central

West Central

Overall

Figure 3.12: Strike Ban Approval, East Central, West Central

the Northeast was negatively correlated with most other regions. However, the aver-

age level of support was above 60 percent in all regions. In terms of class di↵erences,

the unemployed were in opposition to the upper class, but otherwise the classes were

positively correlated. These di↵erences are small in percentage point terms, as each

class favored a defense strike ban by more than 70 percent in each survey. Here we

can see that those who were received public benefits were the least likely to support

a ban on defense strikes; the reserve army of the unemployed was the least likely to

countenance strike bans in support of the war e↵ort.

All regions were in favor of there being a measure adopted to forbid strikes

in defense industries. But the South and Southwest region - the Deep South and

Sunbelt - consistently reported the greatest willingness to see a law forbidding strikes

altogether. This is in keeping with the fact that this region would be the home

to the first right-to-work laws. Within the class subgroups, the relative ordering is

96

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr-41

May-41

Jun-4

1

Jul-4

1

Aug-41

Sep-41

Oct-41

Nov-41

Dec-41

Per

cent

App

rove

Date

SSW

Mountain West

Overall

Figure 3.13: Strike Ban Approval, South and Southwest, Mountain West

consistent with the expectation that the a✏uent would favor more repressive laws but

crucially about three-quarters the poor and unemployed supported such a provision

in the months preceding the attack on Pearl Harbor. It should be noted that for all

the publicity attending the defense strikes in the early 1940s, there was actually less

militance in 1941 than in any of the preceding ten years. Moreover, average levels of

strike volume varied greatly across the regions, in contrast to the general homogeneity

of opinion, as all groups opposed defense industry strikes at a high level.

The questions on the Wagner Act vary slightly in wording but prompt the

respondent to identify as someone who would leave the Wagner Act unchanged or

not. Since the law was evidently a tool for disruptive organizing and a casus belli

for sit-down strikes, opinion on the law reflected sentiment about organized labor

as a powerful force. Labor organizing under the law was the result of the Wagner

Act being adopted and held constitutional, therefore opposition to the law implied

97

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr-41

May-41

Jun-4

1

Jul-4

1

Aug-41

Sep-41

Oct-41

Nov-41

Dec-41

Per

cent

App

rove

Date

Upper

Middle

Overall

Figure 3.14: Strike Ban Approval, Upper and Middle Class

opposition to industrial unionism.

Figure 3.16 shows the level of support for leaving the Wagner Act unchanged in

the Northeast, Mid-Atlantic, and Pacific Coast regions plotted alongside the national

average. These regional opinion moved in tandem with the rest of the country; though,

unsurprisingly, the NLRA was more popular in these core states than the rest of the

country. Unlike the survey items examined above, the Pacific Coast did not support

the pro-CIO response at an elevated level.

In Figure 3.17, we observe a similar pattern within the East Central region as

well as in the West Central region. Perhaps surprisingly, in Figure 18, the South

and Southwest did not markedly di↵er from the national average within and across

surveys; likewise, in Figure 39 the Mountain West tracks the national average.

Across regions, there was never a majority for leaving the Wagner Act un-

98

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr-41

May-41

Jun-4

1

Jul-4

1

Aug-41

Sep-41

Oct-41

Nov-41

Dec-41

Per

cent

App

rove

Date

Lower

Relief

Overall

Figure 3.15: Strike Ban Approval, Lower Class and Unemployed

changed, aside from an outlier reading in the Northeast. Though a majority of citizens

did approve of unions generally, there was never a great deal of popular support for

the law that had legitimized labor militance. Unhappily, though, these surveys began

after the 1937 strikes, so there is no pre-Jones and Laughlin baseline for comparison.

Nonetheless, it would be correct to state that there did not exist a solid pro-Wagner

issue public that over time would have constituted a majority in any region.

Figures 3.19 and 3.20 depict the level of support for leaving the Wagner Act

unchanged by class. Predictably, as shown in Figure 19 the upper-class respondents

were less likely than average to oppose amending the Wagner Act. The middle-class,

being the largest group in these surveys, tracks the national average closely. The

lower-class respondents, as depicted in Figure 3.20, were likelier to support theWagner

Act than average, though, importantly, those survey measures move in parallel with

the national average, nor are they well-elevated above the average. Likewise, in

99

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jul-3

8

Oct

-38

Jan-

39

Apr

-39

Jul-3

9

Oct

-39

Jan-

40

Apr

-40

Jul-4

0

Oct

-40

Per

cent

App

rove

Date

Northeast

Middle Atlantic

Pacific Coast

Total

Figure 3.16: Wagner Act Approval, Northeast, Mid-Atlantic, Pacific

Figure 3.20, we observe that those on the relief rolls were overwhelmingly more likely

to support the Wagner Act; however, even this public moves in a parallel fashion.

Nonetheless, there is some evidence of class cleavage on the Wagner Act, indeed as

the NLRB carried on its work through the 1930s, the lower class and unemployed

respondents formed a majority in support of the law. However, as Gross (1974; 1981)

explains, the Board was changed in a more anti-CIO direction after 1938, so the

machinery of the Wagner Act was changed over the course of the surveys during the

late 1930s.

Regional opinion toward the Wagner Act was largely correlated as all but one

regional correlation was negative. Unsurprisingly, this was the South and Southwest

region, which found the Wagner rights especially destabilizing. Contrary to the sur-

veys examined above, the upper-class respondents and the lower-class respondents

were positively related, though weakly. Nonetheless, the upper-class respondents

100

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jul-3

8

Oct

-38

Jan-

39

Apr

-39

Jul-3

9

Oct

-39

Jan-

40

Apr

-40

Jul-4

0

Oct

-40

Per

cent

App

rove

Date

East Central

West Central

Total

Figure 3.17: Wagner Act Approval, East Central, West Central

were the most weakly correlated with the national average among the class groups.

However, over the course of 1939, support for leaving the Wagner Act unchanged

declined and then rose at the end of the year: 54 percent of respondents would have

left the Wagner Act unamended in February 1939, falling to 34 percent in Novem-

ber that year. The public disfavor toward the Wagner Act coincides with the Smith

Committee hearings regarding charges of unfair application of the Wagner Act by

the National Labor Relations Board. The Smith Committee recommendations drew

union-hampering provisions from the Wisconsin and Pennsylvania 1939 ’Labor Peace’

Acts (Gross 1981, Chapter 4).

On a regional basis there was clear evidence of subgroup parallelism. Unsur-

prisingly, the Northeast held the Wagner Act in the highest esteem while the South

and Mountain West were the least amenable to keeping the law on the books un-

changed. The class subgroups reveal further evidence of parallel publics with respect

101

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jul-3

8 O

ct-3

8 Ja

n-39

A

pr-3

9 Ju

l-39

Oct

-39

Jan-

40

Apr

-40

Jul-4

0 O

ct-4

0

Per

cent

App

rove

Date

South and Southwest

Mountain West

Total

Figure 3.18: Wagner Act Approval, South and Southwest, Mountain West

to Wagner emendation. What is more, the lower status and the unemployed respon-

dents did move in tandem with the other groups. That the unemployed would have

pro-Wagner opinion does not follow unless it is assumed that unions were class ve-

hicles.9 Predictably, the highest status respondents were the least inclined to leave

the law unchanged. For the most part, before the second World War, the public was

already quite open to the idea of amending the Wagner Act and this viewpoint was

not limited to the owners of capital and the managerial class. The strike-hampering

provisions that characterize the present legal regime were adopted by governments

that were in line with mass preference in policy.

Though Roosevelt was lukewarm on the Wagner Act, it came to be a central

legislative achievement, since the unions it legitimated were evidently a major com-

9Which they assuredly were not; not even in the eyes of the most militant leader of them all,John L. Lewis.

102

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jun-

38

Aug

-38

Oct

-38

Dec

-38

Feb-

39

Apr

-39

Jun-

39

Aug

-39

Oct

-39

Dec

-39

Feb-

40

Apr

-40

Jun-

40

Aug

-40

Oct

-40

Per

cent

App

rove

Date

Upper

Middle

Total

Figure 3.19: Wagner Act Approval, Upper and Middle Class

ponent of his landslide reelection; further, the Wagner Act would be the proximate

cause for the court-packing scheme, as the Court was evidently inclined to strike

down the Wagner Act. For this reason, I graph approval of FDR against support for

leaving the Wagner Act unchanged. In this respect, I am asking whether those who

supported the president would have tended to prefer leaving the law he championed

unchanged. Though the public on average expressed 33 percent support for leaving

Wagner unchanged while those who approved of FDR were 45 percent in favor of

leaving the Wagner law unchanged. But here, as shown in Figure 21, the pro-FDR

opinion converged toward the view that Wagner should be amended. And, after the

1938 elections, the president would move to sta↵ the National Labor Relations Board

with members more inimical to the CIO (Gross 1974, Chapter 4).

In 1937, public opinion on the Court packing episode is measured by responses

to the question: Do you approve of the President’s plan to enlarge the Supreme

103

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jul-3

8

Oct

-38

Jan-

39

Apr

-39

Jul-3

9

Oct

-39

Jan-

40

Apr

-40

Jul-4

0

Oct

-40

Per

cent

App

rove

Date

Lower

Unemployed

Total

Figure 3.20: Wagner Act Approval, Lower Class and Unemployed

Court? The subsample responses are reported in Figures 3.22 to 3.24. The regional

subgroups reveal an interesting pattern in that the Pacific Northwest (Figure 3.22)

and the South (Figure 3.24) were the most supportive of enlarging the Court though

all groups converged on the mean by mid-1937. Overall, the public expressed 48

percent support on average for the court-packing scheme, though this was a 52 percent

average before Jones and Laughlin and 44 percent thereafter. The South responded

with the greatest inclination to enlarge the Supreme Court; this is odd in light of

the region’s typical conservatism. On a class basis, the lower class (Figure 3.26)

moved closer toward the average that believed Article 3 of the Constitution should be

left intact, declining from a 75 percent average before the Court reversed tack to 68

percent afterward. Thus, when the constitutional system came under frontal assault

from a perceived class agent - or at least a broad reformer - the mass public favored

maintaining the constitution it inherited. Or, as Caldeira (1987) found in his study

104

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

Apr

-38

Jun-

38

Aug

-38

Oct

-38

Dec

-38

Feb-

39

Apr

-39

Jun-

39

Aug

-39

Oct

-39

Dec

-39

Per

cent

App

rove

Date

Approve FDR

Disapprove FDR

Overall

Figure 3.21: Wagner Act Approval, by FDR approval

of these same Gallup polls, as the Court moderated its position, the mass public that

supported the president lost its appetite for Court expansion. However, as Figures

3.25 and 3.26 show, the upper-class respondents were consistently below the national

average in their approval of the court-packing scheme while those on relief were more

likely to support enlarging the Court than average. Nonetheless, these two groups

moved in tandem with the national average.

The states in New England were negatively correlated with the other regions.

And, indeed, the New England states were opposed to the enlargement plan in each

survey, in contrast to the rest of the nation, which tended to favor the plan before the

Supreme Court upheld Wagner. Most regions averaged around 50 percent support

for Court-packing while the Northeast averaged 39 percent.

The magnitude of change between polls in the time-series are also evidence

105

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

2/1/37 3/1/37 4/1/37 5/1/37 6/1/37

Per

cent

App

rove

Date

Northeast

Middle Atlantic

Pacific

Total

Figure 3.22: Court-packing Approval, Northeast, Mid-Atlantic, Pacific

of opinion change. The regional and class subgroups tended to change about the

same amount between survey readings on leaving the Wagner Act unchanged. A

consistent outlier, however, was the Mountain West, which is attributable to some

combination of measurement error and social polarization arising from strikes in ex-

tractive industries. The unemployed were the most unstable in variation between

measures. A similar story applies to the defense strike items. Again, the Moun-

tain West was the had the greatest variation between poll readings but the classes

were more stable. Variation among the union approval and John L. Lewis data re-

veal an interesting pattern where the upper class was the most unstable, but this is

attributable to rapid increases in disapproval toward unions and their increasingly

expansive demands.

Taken together, with respect to the union approval and labor militance items,

while the socio-economic class groups located themselves roughly in the expected

106

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

2/1/37 3/1/37 4/1/37 5/1/37 6/1/37

Per

cent

App

rove

Date

East Central

West Central

Total

Figure 3.23: Court-packing Approval, East Central, West Central

order, there was often not clear di↵erentiation between the lower classes and the rest.

Any overall coherence in opinion patterns applies to the super aggregated opinion

as well as any group therein.10 The opinion data presented above reflects a mass

public that was largely in sync in terms of levels of support or disfavor toward unions

although some class and regional di↵erences existed. The mass public tended to

disfavor the sit-down method and it tended to hold strikes against the war e↵ort -

even before the war began - in low regard. Indeed, in 1937, in two Gallup polls,

the public favored the adoption of a ban on sit-down strikes at a 60 percent level.

However, the respondents on the relief rolls were evenly split. This is in keeping

with Page, Shapiro and Dempsey’s (1987, 37) observation that the mass public looks

askance on militant protest. This data historicize that result further. But this raises

10It is worth noting that the classes retain their order within each region. The interaction betweenregion and class is small.

107

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

2/1/37

3/1/37

4/1/37

5/1/37

6/1/37

Per

cent

App

rove

Date

South and Southwest

Rocky Mountain

Total

Figure 3.24: Court-packing Approval, South and Southwest, Mountain West

the question of what explains why, say, militant protest against the Vietnam War

succeeded where militant sit-downs in the 1930s failed? In part, the news media was

increasingly friendly to protests as the war wore on (Zaller 1992) - which raises a

question as to why the media moved in one case and not the other. The previous

chapter suggests that editors were more implacably opposed to labor militance than

literal militance.

Substantively, the mass public was in favor of laws that would greatly cir-

cumscribe strikes prior to the beginning of the War, and that the run-up to it only

hardened those views. Furthermore, the public had little regard or sympathy for the

militant union leader, John L. Lewis. This was likely more due to Lewis’ defiance of

the government than of capital. Indeed, it was public favor toward the president that

served to isolate the radical labor leader.

108

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

2/1/37 3/1/37 4/1/37 5/1/37 6/1/37

Per

cent

App

rove

Date

Upper

Middle

Total

Figure 3.25: Court-packing Approval, Upper and Middle Class

The direct causal link between media and opinion is untestable directly given

the sparseness of the opinion data. Were the time-series fuller, then lagged variables

would lend inferential leverage. However, it is implausible to assert that there was

no connection between the homogeneous coverage of strikes and the homogeneity of

national opinion. Moreover, the correlation coe�cients presented below show a weak

relationship between opinion at year t and opinion at year t+ 1 during the 1930s. In

Figure 3.27, the union approval time series is juxtaposed with the volume of strike or

CIO coverage, an inverse relationship is apparent (r=-.03 and r=-.24, respectively).

As coverage increases, negative sentiment toward unions seems to increase as well.

Though not dispositive, this is suggestive and consistent with the argument laid out

in this chapter. The Lewis time series is much shorter, not even the entire Wagner to

Taft-Hartley period, but the same pattern is apparent: the increased focus on Lewis

implied a decrease in the esteem in which he is held by the mass public (r=-.51).

109

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

2/1/37 3/1/37 4/1/37 5/1/37 6/1/37

Per

cent

App

rove

Date

Lower

Relief

Total

Figure 3.26: Court-packing Approval, Lower Class and Unemployed

This is not surprising in light of his well-known opposition to defense mobilization

(Lichtenstein 1998, 45-8). Furthermore, when the count on strikes and CIO articles is

compared with the strike volume measure, a fair connection is evident. Coverage was

tracking the amount of industrial disputes closely on an annual basis. The relative

salience of strikes was not stoked by editorial decisions. Had editors wanted to expose

readers to strike news, then the volume of coverage would have been less sensitive to

the volume of strikes.

Because mass opinion was never clearly on the side of the radical workers - and,

indeed, tended against them - elites never had to take seriously the claims of left-

leaning workers. This, combined with capitalist structural power, served to ensure

that a quick retrenchment of strike laws was a nigh-foregone conclusion. The backlash

was partially ensured by citizens hostile to working-class militance and indi↵erent

to working-class demands (Newman and Jacobs 2010, Chapter 1). Drawing upon

110

0%

20%

40%

60%

80%

100%

0 500

1000 1500 2000 2500 3000 3500 4000 4500 5000

1936

19

37

1938

19

39

1940

19

41

1942

19

43

1944

19

45

1946

19

47

#REF!

#REF!

#REF!

Figure 3.27: Strike Coverage and Union Approval

correspondence to President Roosevelt drawn from their work: ”Most of us would

not accept public gratuities - they are the gift of providence to the profligate and

lazy and foolish for the greater part, though to be sure, all these have votes, and

no doubt in excess of the hard workers and self-respecters” (Ibid, 18). ”You pamper

poverty and throttle thrift. [...] Your administration has coddled those whom it

was politically expedient to conciliate and imposed upon those who, like myself are

politically unorganized although we probably pay more taxes in the aggregate and

have a greater stake in the country than any other class” (Ibid 20-1). ”The middle

classes have su↵ered a great deal [...] What has the government done to help this

class? They do not belong to a union, so nothing has been done but to raise their

taxes” (Ibid, 25). ”The fall of prices took away our farm and savings. In spite of

this, we have managed to send eight of our children through high school [...] We

and our children, trying to keep from relief, are taxed 3 cents on every dollar to

keep these people who have all the things we cannot have and when I complained

to the authorities, they said, why you haven’t been refused relief have you? All of

111

our children work and not through luck. But if they have to give us everything they

are being penalized for being industrious and are paying two sets of relief taxes -

one for us and one for the neighbors who rest while the children work. We object

to having relievers trying to put us on relief” (Ibid, 26-7). ”As it is carried on, [the

relief program] is making paupers faster than any method ever devised [...] The relief

families are sure learning fast how to work the racket and live better than the average

farmer” (Ibid, 26). This correspondence to the White House from ordinary people

who were often skeptical of government assistance exemplified why anti-strike laws

more likely to be supported than opposed.

What, in the end, seems to have doomed working class radicalism in the United

States is the lack of class consciousness on the part of its citizens. A deeper analysis

might have been possible were researchers able to supplement the survey responses

with just a bit more information about the respondents - vote choice, for example

- a more comprehensive analysis would be possible. However, in terms of providing

relief or allowing for legitimate worker protest, there seems to have been little e↵ect

from class consciousness. People who might have seen themselves as being deprived

of the value they had created instead seem to have blamed their lot on themselves or

Providence, being mystified by capitalism.

The meaning and content of strikes changed over the course of the New Deal.

Sit-down strikes became defense strikes which became Reconversion strikes. The

strikes were varied in their ends and sit-down strikes in defiance of statutes became

less common over time (Lichtenstein 2013, Chapter 2). What unified the strikes, im-

portantly, was that the strikes were costly to production. The initial wave of sit-down

strikes were aimed at recognition while the defense strikes were toward multiple and

di↵use ends while the reconversion strikes - which are typically viewed as a proximate

112

cause for the adoption of Taft-Hartley - were attempts at forging workplace citizen-

ship (see Lambert 2005, Chapter 5 for a discussion of UAW- and CIO-led proposals).

Likewise, temporal variation in opinion regarding strikes was not substantial and was

generally negative. Nonetheless, the citizens’ divide between work and private life is

bounded by a constellation of psychological factors that are historically constructed

and attenuate the e↵ect of class on citizens’ worldview (Katznelson 1981, 6).

This analysis is consistent with the Schickler and Caughey (2011, 186) conclu-

sion that while ”public views on labor unions in general were consistently favorable,

attitudes toward pro-union policies, aggressive labor tactics, and militant labor lead-

ers were highly negative.” However, in contrast to that work where the authors looked

more closely at the 1940s, here this conclusion is shown to be applicable to the US

prior to Pearl Harbor and, indeed, prior to the 1938 elections. By turning the focus

onto the critical year of 1937-8, we get snapshots of the contours of mass opinion,

which played heavily into the considerations of elites (Brinkley 1995; Eisinger and

Brown 1998). Since the mass public and elites were comfortable with the existence of

unions while disfavoring the militance directed toward programmatic ends, retrench-

ment was an unavoidable result.

As Page and Shapiro (1992, 287-8) point out, ”public opinion may change in a

uniform fashion across the whole population, if the national media transmit similar

information to all groups and the receive and process it in similar ways. This is

especially likely to be true to the extent that di↵erent groups have similar interest or

think in terms of public good and the collective, or national interest.” The collective

interest is not a given; it is constructed in large part by elites. An important, though

soon to be supplanted, medium for the transmission of pro-capital messages were

newspapers. Elites sought to shape opinion and industrial unionists were hard done

113

by it. And, as Fones-Wolf (2000) points out, the CIO unions were aware of the

media slant against them and recognized it as an obstacle to organizing and coalition

formation.

Bearing the state cases in mind, it is worth emphasizing that the regional dif-

ferences in opinion were relatively muted across surveys, each region was typically

positively correlated with all other regions. Despite the distinctiveness of the local

editors, the overall patterns of approval were consistent. Therefore, citizens in dif-

ferent states and regions cannot be said to have been marked by significant opinion

cleavages. And, at the state level, the governing parties in New York and Wisconsin

made similar moves. The Wisconsin state Republican Party in 1935 ”pledged the

worker the right to singly, or collectively bargain for working conditions and wages,

either as an individual or through a representative of an organization of his own

choosing,” but in 1937, the GOP ”pledged to defend the right of labor to organize

and bargain collectively” (Blue Book 1935, 487; Blue Book 1937, 282). The New York

Democrats expanded their definitions of unfair labor practices under the protective

regime to include the unfair practices of labor unions (Red Book 1939, 146). In Utah,

the recognition of collective bargaining was dependent, in part, with the assent of the

Mormon leadership, in light of the perceived communist threat posed by mine worker

organizers in competition with Lewis’ outfit (Powell 1977, 98-100). Utah, though,

had a labor board that was far less busy than its counterparts in New York and Wis-

consin, handling less than ten percent of the caseload of the other two states (Cohen

1948, 11). In other words, despite the fact that strikes were di↵erentially disruptive

across political units, the attitudes toward them were similar.

The above presentation of regional sub-groups strongly suggest that the states

that adopted Wagner-style laws were not marked by notable divergence in opinion

114

toward unions or strikers. The various regions moved in tandem and had similar

levels of approval across survey items. The absence of state-level opinion data is

lamentable, but it does not undermine the conclusion that regional parallelism was

the rule with regard to worker protest before, during, and after the Second World

War. Indeed, it would be fairly unlikely that the states of New York, Wisconsin, and

Utah were all substantial outliers from their neighbors. Likewise, though there were

di↵erences in levels of approval by class, there was no divergent movement between

the classes; that is, the sub-groups moved in tandem even though those on relief and

the upper-class were distinct. The absence of contrary sub-group dynamics evidences

the lack of class consciousness at the height of labor’s protest power.

Conclusion

In the first chapter, I noted that the same militance that won statutory gains

for labor was the same militance that ensured statutory losses. In this chapter and

the previous one, I asked whether that same militance had di↵erent results because

of media presentation of strikes. Perhaps, though, there exists a reason for a sub-

stantial change in public opinion and voting behavior in the changing nature of strike

grievances. The strikes that won the Wagner Act were strikes for recognition whereas

the strikes that followed were strikes in furtherance of union wage and benefit de-

mands, or in sympathy with strikes of that nature. Therefore, the mass public cannot

be said to have been manipulated. In broad strokes, the policy outcomes people got

were in line with the policy outcomes they received.

It would have been doubly tragic if the inability of the press to realize its

democratic ideal was instrumental to the reversal of industrial democracy. This was

not the case. Though the newspapers did emphasize the social costs of strikes without

contextualizing of the worker grievances, the mass public cannot be said to have held

115

the opinion that the labor policy should have been more to the left. And, indeed, there

is clear evidence that citizens did not wish to tolerate the industrial action workers

had to take in order to maintain bargaining strength against employers.

The backlash against unions played out in the press though the mass public

was clearly disposed toward taking the view that strikes to ensure union bargaining

power were not worth the social cost. And, as we shall see in the next chapter, the

strikes that prevented militant unions from garnering public sympathy were strikes

that prevented labor from being part of a winning legislative coalition. The labor

stoppages deprived consumers of desired goods while antagonizing essential legislative

partners.

Citizens evidently had a limited respect for unions. Though happy to see unions

exist, the mass public did not support radical militance nor did they support strikes

in conflict with the war e↵ort, prior to Pearl Harbor. Furthermore, the public did

not support the Roosevelt presidency to the extent that lack of resolve on the part

of FDR can be blamed for the delegitimizing of worker resistance. On the contrary,

citizens supported FDR’s court packing until the Court agreed to uphold the collective

bargaining law - note the decline in support across all groups after the April 12, 1937

’switch in time.’

The mass public accepted that unions should exist but generally had little

patience for labor militance. In light of this, the failure of congressional conservatives

to enact more repressive retrenchment programs can be viewed as a victory of sorts

for the labor movement. Indeed, this dissertation has wrestled with the question of

whether to treat the development of American labor relations as a glass that is half-

full or half-empty. However, an interpretation that laments the missed opportunities

of the New Deal is undermined with reference to the contours of mass opinion. We

116

are getting close to definitively ruling out the ’half-empty’ perspective. In the next

chapter, we will do so.

117

Chapter 4

State-level Origins of Anti-Strike

Laws

Introduction

The introductory chapter presented the states as usable sites for comparative

investigation during the Wagner to Taft-Hartley interlude. The succeeding two chap-

ters framed the information environment and opinion patterns in which the states

were situated. The citizens were found to be generally opposed to strikes and to

have been presented newspaper articles that were generally unfavorable to strikes

and strikers. Given that the various states shared a common discourse that was

unfavorable to radical worker protest, in this chapter I ask, why did states adopt

anti-strike laws? Further, why did states that adopted Wagner copycats adopt re-

trenching amendments in short order? This chapter will use a series of regressions to

address the first question and reference to our key state cases to address the second;

in both elements, the variable of interest is whether a state adopted an anti-strike

law in a given year.

118

Recalling the introduction, this chapter assumes that the states are su�ciently

similar to invite comparison. The analysis in this chapter also assumes that the pol-

itics within the states is su�ciently similar with the federal government to invite

comparison. From an inferential perspective, broader comparison enables broader

generalization about the logics governing politics. These assumptions are not partic-

ularly heroic because strikes lay bare much of the political struggle that organized

politics seeks to sublimate.

Implicit in the preceding analysis is the assumption that the object of political

contestation in the various political units is the opinion of the mass public. That is,

the struggle for chamber majorities and friendly Labor Boards called for the mar-

shaling of favorable opinions. And, from the late 1930s into the WWII period, public

opinion turned against unions and this fact did not escape the notice of labor (Foster

1975, Chapters 1-2) or capital (Fones-Wolf 1994, Chapter 2).

The questions this chapter addresses also bear on the counterfactual wherein

Wagner is left untouched. In particular, the cases exemplify anti-union political

dynamics and suggest why alternative pro-union paths were occluded. In order for

the unionists to have defended their early-Wagner strike rights, the labor movement

would have had to organize and strike in a fashion that was costly to owners without

spillover costs being absorbed by potential coalition partners. In practice, chief among

these partners was farmers, who were quick to turn against the CIO. The potential

for alternate coalition partners was dim at the time - quite rightly, the CIO sought

to form an alliance with the NAACP in the following decade (cf. Baylor 2013). But,

the third party option was evidently precluded in the 1930s, in light of the experience

of the states.

The logic that produced the speedy retrenchment against organized labor evi-

119

dently applied to states that adopted Taft-Hartley precursor laws as well as the nation

as a whole. Most important among these provisions were those pertaining to strikes

because those laws served to turn the coercive apparatus against disruptive worker

resistance. To begin, I aggregate the restrictive laws adopted by states pertaining

to strikes that were subsequently in the Taft-Hartley law. In the remainder of this

chapter, I will investigate the cases of several states that adopted protective laws

after Wagner was upheld in order to draw out the logic of the previous chapters and

to address the question of why the disruptive militance of the New Deal period was

double-edged.

The cases will treat three of the five states that adopted protective laws mod-

eled after the Wagner Act: New York, Massachusetts, Pennsylvania, Wisconsin, and

Utah. I leave Massachusetts o↵ to the side because its protective law provided for

mediation but not for a Labor Board; therefore the state administrative apparatus

was not tightly coupled with the militant industrial unions as in the other cases.

The other four laws established a Board with adjudicatory powers regarding labor

disputes. The Pennsylvania case will be referenced but is not a central focus because

the labor stoppages in that state were marked by protracted and bloody showdowns

(Lichtenstein 2012; Davin 2000). That is, the disruption in Pennsylvania, particu-

larly in the ’little steel’ conflicts, manifested the kind of violent class opposition that

democratic politics is meant to sublimate. The remaining three states vary over the

outcome of interest but are su�ciently similar as to warrant comparison. And, we

shall see that the seeming ’odd duck’ of Utah serves to elucidate the logic as an

exceptions which proves the rule of disruption-backlash retrenchment.

The New Deal period was crucial in the development of the American state.

And the expansion of the labor relations bureaucracy was one of the defining fea-

120

tures of the New Deal program. The statutory changes in Norris-LaGuardia (the

anti-injunction Act) and Wagner (the National Labor Relations Act) spurred new or-

ganizing, particularly in auto, steel, rubber, and garment unions (Zieger 1986, Chap

2). When these laws were adopted, the industrial unions - the CIO unions - grew.1

And when the CIO grew, as we shall see, it caused the fracture of the political coali-

tions that adopted those laws in the first place.

Recent studies in APD on the origins of fair employment legislation in the

states and the alliance between the industrial union and civil rights movements that

realigned the parties have used proxies for CIO power, Republican control of veto

points, and state-level variation in racial makeup as variables of interest (Zeitlin and

Weyher 2003; Chen 2007; Baylor 2013, for example). In the following I investigate in

a similar fashion, using state-years as my unit of analysis, bearing in mind that the

restrictive statutes were reflective of mediated public opinion.

It is unlikely the protective statutes would have been passed at the state level

in any instances had the federal government not passed the Wagner Act first. The

construction of the Labor Boards with subpoena powers was necessary to give the

pro-union laws force. And, it was the organizing spurred by the Boards and the

very passage of the Wagner Act that caused the restrictive majorities to coalesce.

A related question is why do states without protective laws adopt restrictive ones

subsequently? That is, outside the South, is the policy shift a story of Republican

seizure of legislative control spurred by militant union disruption? If so, then, this

pattern of protection to restriction is just a story of capitalist disarray and subsequent

reorganization. But an argument that places capitalist elites at the center of the story

needs an explanation for the widespread application of social democratic policies in

1Sadly, there are not data on state level membership.

121

other countries. Although the American businessman has a unique distrust of his

state (Vogel 1978; 1983), capital’s interest with respect to labor is invariant across

time and space. More subtly, there was cross-national experience with social insurance

policies in the 1930s (Amenta et al. 2001, 224); an argument that emphasizes the

decisive impact of capital in the content of the Wagner Act needs to explain why

elites who shared information arrived at wildly varying policies despite having the

same interest.

Substantively, the protective laws entailed the formation of new institutions -

the Labor Boards - to expand labor’s capacity. This was evidently the understanding

that the CIO unions had, for every industry covered was faced with organizing strikes

(Foster 1975, 8-11). And crucial to the explosive growth of the then-renegade AFL

unions was the belief on their part that the Wagner Act had created a new space for

legitimate, disruptive action. The Boards were the crucial enforcement mechanism

for the protective labor relations acts.

The political undoing of the National Labor Relations Board has been treated

extensively elsewhere (Gross 1974; 1981). However, the institutional irony of the

labor relations acts is that their very energy was the primary cause of their undoing.

The power of the young CIO within the Democratic Party was dependent on the legal

protection and relief provided by a Labor Board. Because the new labor boards were

active proponents of CIO organization, they served to work as an opposing force to

the AFL which increased the AFL’s willingness to join the Republican coalition in

some states, such as New York, as noted below.

In theory, a backlash implies that there existed a previous equilibrium that had

been disturbed (Pierson 2000, 85). Before the Wagner Act, would-be union militants

had little reason to expect that statutory gains would be e↵ective in the face of

122

judicial opposition. The Supreme Court upholding Wagner in the face of political

pressure solidified industrial union gains. The newly legitimized industrial unions

were wedded to the Democratic Party, because the Labor Relations Boards that made

the CIO growth possible were reliant on Democratic control. In e↵ect, this served to

contrapose unions with preexisting members of (state) Democratic coalitions, be they

AFL-allied machines, agricultural interests, or white supremacists (Fink 1969; also

discussions of Wisconsin Farmer-Labor tensions throughout the dissertation speak to

this).

What this contraposing chiefly entailed was the CIO threatening to make the

AFL irrelevant; or to displace agriculture interests. Recalling the discussion of feed-

back and equilibria in the first chapter, what CIO organizing strikes served to do was

disturb an existing equilibrium. Labor was strong enough to break the old arrange-

ment but it had far too little leverage to determine the emerging equilibrium.

There was a shift in the legitimate exercise of authority in labor relations as

a result of the Wagner Act. New institutional actors, in the form of the industrial

unions and the Boards, carved out space at the expense of the AFL, farmers, and the

urban machines. What happened at the state level was not unlike what happened at

the national level: widespread and salient, strikes served to generate public disfavor

that formed the basis for Republican majorities (outside the South). The result was a

fracturing of Democratic coalitions at the sub-national level, which led to Republican

seat gains. The 1938 backlash was the result of union militancy under the Wagner

regime.

Did strikers empowered by protective labor laws cause the adoption of restrictive

laws? In order to examine the extent to which the concomitant backlash to CIO

growth was determinative over restrictive statute adoption, I outline the analysis

123

in four parts below: first, , a description of the data under investigation and the

method; second, the results, which o↵er modest support, under a stern test of the

theory; third, the case studies; finally, I conclude with thoughts on further study.

This investigation is in the service of amending the literature on the role of labor

in politics in the New Deal and the decline of US private-sector unions to include a

place for outcomes at the state-level and to leverage these state-level comparisons to

highlight the logics seeming to govern the political conflict between labor and capital.

Statutory prohibitions on sit-down strikes, sympathy strikes, ’hot cargo’ strikes, and

jurisdictional strikes all served to diminish the range of legitimate disruptive activity

in which unions, particularly industrial ones, might engage.

The environment of state politics was embedded in a context that had been

defined by their failure to reverse the setbacks of the Depression (Patterson 1969

was among the first to notice this). In response to the shortfalls, the federal govern-

ment expanded its spending and administrative capacity. Whether state government

political and policy outcomes were more a function of partisanship or economic cir-

cumstance is an open question on the face of it. The economic crisis was deep and

the 1938 recession approached the nadir of the Roosevelt years.

[TABLE 4.1 ABOUT HERE]

Ostensibly, the strikers made great gains in their organizational membership

(see Table 4.1) through recognition strikes in the core industries and would have been

well-served to maintain their independence from the Democratic administration. And,

on the face of it, organized labor’s access to FDR’s burgeoning administrative state

was the basis for a relationship worth cultivating. However, Roosevelt’s selective

granting of access to Hillman and Murray (accomodationist leaders of CIO unions in

textiles and steel, respectively) over Lewis (the militant head of the CIO and the mine

124

workers) helped deepen the rift within the industrial labor movement and wed the

mainstream of the CIO to the Democratic Party (Lichtenstein 1982, Chap. 3; Milton

1982, Chap 4). Rather than seeking their own fortunes under a separate banner as

Lewis recommended, the CIO threw their lot in with the Roosevelt administration

and with the Democratic Party outside the South.

This is not to suggest that the marriage with the Democratic Party was an

error, as such. It is hard to argue that Hillman was not as much FDR’s envoy to

labor as he was labor’s envoy to the President. However, alternate coalitions were

tried at the state level, and courses that led unions to be aloof from the Democrats also

failed to forestall retrenchment. Put all together, the hard left proposition that labor

might have secured a more favorable outcome through adherence to labor stoppages

is di�cult to sustain. More subtly, access to government is of varying and quite

limited degrees of usefulness to a labor movement that cannot credibly countenance

voluntarism. Voluntarism is the idea that unions should ’reward friends and punish

enemies’ as Gompers said - it is the idea that labor should support whichever choice

in a given election that is more palatable, or preferable. But this idea is inapplicable

in a context in which one party is highly oppositional.

Outside the South, we will see that the states that adopted union restriction

laws, with a broad or narrow statutory scope, were heterogeneous in their partisan

composition as well as factors of production. Within regions there was also significant

variation in statutory outcomes. As noted above, New York and Pennsylvania had

far di↵erent labor relations experiences (as did Oregon and California or Indiana and

Ohio). Neighboring states with similar characteristics traveled di↵erent paths.

This analysis of restrictive statute adoption leaves the committees in the cham-

bers unobserved. As the Ways and Means and Defense Committees were essential

125

to Dixiecrat New Deal retrenchment, so were subnational majorities to state-level

retrenchment. However, the data are clearly in support of the claim that there was a

relationship between militance in a state-year and whether a restrictive labor law was

adopted. Although this analysis does not dive into the battles over statute adoption

in the chambers, it is evident that the laws were enacted in response to strikes with an

eye toward curtailing them. This chapter will deploy a mixture of methods to provide

an explanation for the adoption of these statutes. First, the backlash explanation will

be investigated statistically; then, narratively.

Method and Data

Following the periodization in the first chapter, the 48 states between the years

of 1937 and 1947 inclusive, comprise my units of analysis. My outcome of interest is

whether a state-year saw the adoption of a restrictive labor law and my key variables

of interest are strike volume in the state-year and Republican control of a state veto

point, the expectation being that these variables will vary directly with anti-strike law

adoption. Republican control of a veto point connotes whether a legislative chamber

or the governorship was controlled by the Republican Party (Council of State Gov-

ernments, 1937; 1939; 1941; 1943; 1945; 1948). As the Northern Democrats were

motivated by an interest in leveraging the New Deal labor relations institutions to

sustain the growing industrial organized and immigrant voting bases, the Republican

o�ceholders were motivated by the demands of business organizations and agricul-

tural concerns. I do not double-count multiple anti-strike statutes because any and

all of the banned labor stoppages were and remain essential for wage earners to obtain

leverage over their paymasters. I do not code for pro-union statues because none were

adopted aside from the Wagner copycats which will be dealt with in the penultimate

section of the chapter. (See Table 4.8 for exemplar laws. As noted in the first chapter,

126

these laws were primarily bans on sit-down strikes and jurisdictional strikes.)

The disruption that arises from labor stoppages can be measured not only by the

sheer number of workers involved, but also by its duration as well by their frequency.

To this end, following Hibbs (1973), I employ a strike volume measure that accounts

for all three of these factors. This strike volume variable is a product of the size,

duration, and frequency of strikes. Size is the ratio of strikers to number of strikes.

The duration is the ratio of eight-hour days (man-days) lost per striker. The frequency

is the ratio of strikes to 1,000 workers in the labor force. This is called strike volume

because it is akin to the geometric concept of volume: the product of length, height,

and breadth (Ibid, 1036). The strike volume reduces to the ratio of man-days lost

per 1,000 workers. This measure serves as an index of industrial conflict that is more

comprehensive than the count of strikers alone.

Additionally, I include an indicator for whether a state is in the South2 given

the distinctiveness of the states that mandated segregation. Because of the Dixiecrat

interest in preserving racial segmentation, the partisan cleavage over labor relations

is null in those states.

Also importantly, immigrants were politically distinct from the native-born:

the native-born were typically in the skilled AFL while the recent immigrants were

working in the unskilled jobs the CIO sought to organize. And, what is more, the

black population was not overwhelmingly likely to support the Democratic candidate

at the outset of the New Deal, thus helping install a partisan barrier between them

and the white recent immigrants in their economic class.

2The South are taken to be: Alabama, Arkansas, Delaware, Florida, Georgia, Louisiana, Ken-tucky, Maryland, Mississippi, Missouri, North Carolina, Oklahoma, South Carolina, Tennessee,Texas, Virginia, and West Virginia.

127

I attained the proportions of black, immigrant, and urban residents as well as

state income per capita from the Statistical Abstract of the United States and my

expectations are directly informed by the theories of disruption-backlash and labor

market segmentation. A state with more urban residents has more individuals in

a geographic space to be inconvenienced by a strike. A state with more blacks is

likelier to have a more segmented labor force, implying an increased likelihood of

restriction. Finally, state income is a proxy for state capacity; I argue that states

with greater fiscal capacity would be less interested in restricting labor because the

New Deal party within the state has more reason to ambitiously form a ’baby’ New

Deal coalition.

Penalized maximum likelihood estimation can be generally applied to data that

su↵er from quasi-separation to generate unbiased finite estimates (Firth 1993; Heinze

and Schemper 2002). This has the attractive feature of enabling researchers to avoid

dropping covariates or groups (Zorn 2005). That the data su↵er from quasi-separation

or ’sluggish’ indicators is itself a fault of the data. (Bartels 2008, 28). However, the

data o↵er the opportunity to present a challenging test to the hypothesis that strike

disruption was decisive. Because I focus on the adoption of anti-strike laws, the

groups by state and year would be unbalanced if I deployed a standard logit model;

therefore, I use penalized maximum likelihood estimation to find odds-ratios that

would be analogous to those estimated by a standard logistic function.

Therefore, I estimate a logit model of the functional form: Pr(Yik

= 1|Xik

) =

11+exp(�

P�ikXik)

, where Y

ik

is the probability that state i adopts a restrictive law

in year k. X is a vector of indicator variables associated with each state-year.3

3One might want to deploy a Cox regression or some other Event History method (like theseminal Berry and Berry 1994) but the controls in a hazard regression in this dataset would violatethe proportional hazards assumption, and, therefore, bias the models. And, just as importantly,

128

These variables are whether the amount of strikers in a state, Black population,

urban population, State income, as well as indicators for Republican majorities or

governorships and Jim Crow laws.

Results and Discussion

[TABLES 4.2 AND 4.3 ABOUT HERE]

Strike volume finds support as a variable of interest. The strike volume variable,

though significantly associated with restrictive law adoption, appears quite small in

magnitude. By way of interpretation, every increase by one thousand man-days lost

increases the odds of restrictive law adoption by a tenth of a percent; that is, every

extra 10,000 man-days lost increases the probability of restrictive law adoption by one

percent. The e↵ect of this variable is significant alongside statehouse partisanship.

States under at least partial Republican control were about three times as likely

to adopt restrictive laws. The controls for state income, black population, and the

Jim Crow dummy variable are all insignificant, though they carry the expected sign.

Southern states and states with more blacks were more likely to adopt restrictive laws.

States that were relatively more prosperous were more likely to adopt restrictions on

strikes. Finally, the control for urban population is also significant in this model; on

average an increase of state urban population by one percent implies a two percent

decrease in the likelihood of adopting restrictive law.

Turning to the strike volume results for the non-South states in Table 4.3, we see

that the strike volume variable holds the same value and reaches conventional levels

of significance. The GOP veto variable, however, increases greatly in magnitude

estimating a fixed-e↵ects logistic model would entail substantial losses of statistical power as thedata are quasi-separated both temporally and spatially - that is, there exist years and states thathave no variation on the variable of interest. Because of this, variables or groups must be droppedfor a finite solution to arise from optimization.

129

though is estimated less precisely than in the case of the previous regression. Here

an increase in 10,000 man hours implies a one percent increase in the likelihood

of adopting restrictive laws while states with at least partial Republican control are

more than 12.5 times as likely to adopt a restrictive law. In addition, the state income

variable is significant. Since this variable has been logarithmically transformed, the

interpretation is not straightforward: the key takeaway is that increasing state income

implied an increased likelihood of restrictive law adoption. And again, the urban

share of the population is negatively associated with the likelihood of anti-strike law

adoption, an increase of one percent urban resident implied a 3.5 percent decrease in

the likelihood of law adoption.

Though the magnitudes are small for the strike volume variable, the association

is robust to controls and the overall fit is high across models. Each component of the

strike volume variable captures an element of the disruption caused by strikes.

[TABLE 4.4 ABOUT HERE]

The marginal e↵ects for the full model with all forty-eight states are presented

as di↵erences in odds in Table 4.4. In the case of the binary variables, the marginal

e↵ect is that of going from zero to one and those are easily interpreted. The marginal

e↵ects imply probability changes for given values of the variables of interest. In this

case, the probabilities implied by the marginal e↵ects suggests that for every 10,000

man hours lost in a given year, the state was .29 times as likely to adopt a restrictive

law; while moving from no Republican control to at least partial control implied the

state was 1.32 times as likely to experience restrictive law adoption. These results

are consistent with the view that the adoption of restrictive laws was driven by a

backlash to strike action. The other marginals carry the expected signs.

130

The chief impact estimated has been a mediated one: that of strike disruption

producing Republican gains outside the South. What happened in the states occurred

before Taft-Hartley - moreover, the state level retrenchment even occurred before the

1942 Smith-Connally Act was passed. More importantly, the phenomena of speedy

retrenchment and of restrictive law adoption in the states would suggest that the

antipathy to labor union power was broadly shared in the mass polity. That is, the

story of the crystallizing of US labor relations law was essentially democratic.

In this dissertation, CIO militance is proposed to be a neglected explanatory

factor in the literature as a result of the neglect of the states during the New Deal.

But this chapter does not have a direct measure of CIO militance - rather, it proceeds

on the basis of overall militance. Measurement in this fashion runs the risk of conflat-

ing non-CIO militance with CIO-driven outcomes. However, we know from historical

study that the CIO was a major component and instigator of workplace resistance.

Taken together, however, the point estimates tend to support the thesis that restric-

tive law adoption was driven by strike disruption mediated by partisanship.

In the non-Jim Crow states, the Republican electoral victories of 1938 were

themselves the product of disruption wrought by strikes. Or, at a minimum, the

election of conservative legislators is a clear predictor of restrictive law adoption.

This seems to imply a counterfactual that is consistent with the literature: no decline

in business power, no protective laws; no increase in business power, no restrictive

laws (Hacker and Pierson 2002, 318). Yet, to directly speak to this, a uniform proxy

for business strength in the states would need to be deployed. Including a proxy for

a neighboring state having a restrictive law does not add explanatory power to the

analysis. In fact, the measure is highly correlated with southern statehood.

But the preceding analysis raises the question of why Republican legislative

131

majorities came to power at the state level; therefore, I estimate a regression model

of Republican seat gains in the 1938 election.

I draw the Republican seat gain data from the Book of the States (Council of

State Governments 1937; 1939). I add the seats in either chamber and use the sum. I

attained the state income data from the Statistical Abstract of the United States (US

Census Bureau 1940). I attained the FDR 1936 vote share from the California-Santa

Barbara American Presidency Project. Finally, the measure of CIO density is drawn

from the Troy count of dues paying 1939 members. In terms of expectations, as noted

above, CIO density should be associated with Republican seat gains in 1938; state

income4 should vary with Republican seat gains; and, FDR 1936 vote share should

vary directly with Republican seat gains, under a balancing theory (as in Bafumi,

Erikson, and Wlezien 2010).

[TABLE 4.5-4.7 ABOUT HERE]

Further work is needed to evaluate the seeming potential for omitted variable

bias in state labor retrenchment models. Another concern is whether the results in

the tables above were temporally as well as regionally driven. Indeed, it should be

noted that a Chow test performed at the break date of 1940 based on the model in

Table 4.1 yields a significant F�score (see Table 4.5). This indicates that there is a

break in the coe�cients between pre- and post-1939 groups (see Table 4.6 for odds-

ratios grouped around the break date). Put another way, the politics of retrenchment

against worker protest were deeply-seated in US politics. The coe�cients for strike

volume are similar, though only significant in the pre-1939 specification. The value

of Republican partisanship is much higher after 1939, but this coe�cient did not

4Ideally, I would be able to employ a measure for annual change in state income, but the StatisticalAbstract does not allow for this.

132

approach conventional levels of significance. Importantly, 1939 is the only year where

an F -test indicates a break in the regression coe�cients.

The OLS estimates5 for the non-Southern states presented in Table 4.7 show

a strong association between CIO density and Republican seat gains. A one-unit

increase in the CIO density implies an average of 62 additional seats in the state

legislature for the Republicans. That is, adding 1 to the ratio of CIO workers to the

labor force yields the coe�cient estimates. Therefore, each increase in CIO density by

1.5 percent yielded an average increase of one Republican seat. The control variables

are not clearly associated with the GOP gains, which is furthermore consistent with

the understanding that labor militance was a highly salient issue. Nonetheless, they

carry the expected signs here where FDR 1936 vote, state income, and state urbanism

vary inversely with Republican gains, if tenuously. It is particularly important that

the income variable is insignificant because the ’Roosevelt recession’ was as salient

as the widespread labor stoppages. Density, though indirectly, captures the e↵ect of

militance because we know that the CIO could not and did not organize without a

large element of protest. A competing explanation for seat losses, which we engaged

with in the previous chapter, is the court-packing controversy, which resulted in

a salient defeat for FDR. And, importantly, no interaction terms would approach

conventional levels of statistical significance.

Taken together the estimations tabulated above point to the interrelation be-

tween strike disruption conceived narrowly or more broadly, partisan control, and re-

strictive law adoption. This lens points my investigation toward and under-explored

path. The operant sections of the Taft-Hartley Act that circumscribe worker protest

5The N is 29 because Nebraska and Minnesota are omitted since those states had non-partisanlegislatures. However, it is more likely that the results would be strengthened by their inclusion, asthe o�cials who won behaved in a decidedly conservative fashion

133

percolated upwards from these state laws. The experience of the states was a priority

in the drafting and advocacy for the Taft-Hartley Act. For example, as Representative

Hartley (R-NJ) argued: ”In many instances where secondary boycotts have occurred,

union trucks have refused to haul products of honest hard-working American farm-

ers. If the farmers or dairymen try to haul for themselves, their trucks are destroyed

and their lives threatened. For instance, the lettuce strike in Salinas [California] in

1936 caused a loss of 2000 [train] cars of lettuce which would have been distributed

over the United States” (Legislative History, 583). More to the point, Hartley elab-

orated: ”One of the main objectives of this bill is to break down authority to call

strikes which can cripple our entire economy” (Ibid, 784) Furthermore, the legislative

history of the Taft-Hartley Act contains several references to state labor peace laws

and how the national law is meant to be consistent with those statutes (Ibid, 422-4;

881-83, for example). Illustrative of this theme, Taft argued that ”This necessary

measure [anti-picketing provision] is practically contained in the Wisconsin Labor

Relations Act and has been used in Wisconsin as a means not only of preventing the

coercion of employees but also of attempting, bringing such action as can be brought

by administrative law to end mass picketing...It would be the simplest matter and

the matter which will bring up the least litigation to use the same words in this law

[Taft-Hartley]” (Legislative History, 1032-3).

This project does not mean to suggest that the labor laws are mere sca↵olding

for subsequent bargaining games, nor does this study assume the e�cacy of labor

law; indeed, labor law is constitutive of the strategic context shared by the actors.

As Hattam (1990) and Forbath (1991) argue, the (hostile) judicial regulation of the

labor movement necessarily reduced the expectation that programmatic gains in the

legislature would be e↵ectual. Thus, voluntarism and dual unionism were rational

134

results. Indeed, the new labor laws passed in the political units constituted the

grounds for subsequent conflict, particularly in the labor relations boards. Most

importantly, though, the laws that were passed had been designed to control labor

militance. Furthermore, the primacy of militance is evidenced by the fact that New

Deal policy gains were won by the CIO unions before they were enforced by the federal

government. In a sense, the strikes conducted by the nascent CIO unions supplanted

the preexisting political authority which almost certainly contributed to the adoption

of restrictive laws.

Does the preceding mean that the people got the labor laws they wanted?

Seemingly, the unpopular disruption was met with an electoral backlash, itself con-

sonant with opinion polls regarding sit-down strikes and defense strikes presented in

the previous chapter. Restrictions on private-sector union strikes that emerged from

the Wagner to Taft-Hartley period decisively ensured that the principals of corpora-

tions would have an outsized influence on American politics (more on this in the next

chapter).

Cases

In this chapter, the dissertation turned its focus from the media and public opin-

ion to the politics leading up to laws restricting the right to strike. The previous two

chapters have emphasized that the citizens of the various states were, broadly speak-

ing, exposed to the same information and that region did not apparently a↵ect the

way that information was processed. Bearing in mind that regional opinion was fairly

homogeneous, this chapter will be concerned with tracing whether and why states

that adopted Wagner-style laws adopted retrenching acts shortly thereafter.

This dissertation has been attempting to marry the behavioral and historical in-

135

stitutional approaches. Where behavioralism takes actions and expressions to be the

site of revealed preference, historical institutionalism is concerned with how prefer-

ences are constituted and configured at particular times as products, not just causes;

but as causes as well as products.

In the states we will examine below, we will observe that the strikes for recog-

nition that precipitated Wagner reverberated through the political economy. The

rapid success of the conservative backlash within heterogeneous contexts presents the

opportunity for a most-similar comparison. Generalizing through the method of dif-

ference will suggest that the rules of the electoral game largely explain the successful

retrenchment against class protest during the New Deal. Within the two-party struc-

ture the organized labor movement lacked - necessarily so - the leverage to turn the

Democratic Party toward a labor-left agenda.

There are more potential explanations for retrenchment than cases in this chap-

ter. (See Table 4.11 reproduced from the Introduction.) Therefore, I am relying on

the previous chapters to eliminate plausible alternatives from contention. What I am

aiming to draw from the ’laboratory of history’ is a closer causal tracing of the mech-

anisms that governed whether the backlash manifested. And I rely on the similarity

of the states to enable semi-experimental comparison.

Why did states retrench to anti-strike laws? The Wagner Act and the Taft-

Hartley amendments have analogues at the state level during the New Deal. The

adoption of the ’little Wagner Acts’ mirrored the adoption of the national bill and

the adoption of conservative amendments are harbingers of Taft-Hartley; these are

a class of like phenomena that call for explanation. The union security laws created

statutory protection for collective bargaining, thus strengthening the hand of labor

against capital.

136

An important assumption - also relied upon above - is that the 1930s constituted

a decisive break with the past, particularly in the area of labor relations. Though

many proposals were considered by the legislatures early in the twentieth century,

the economic crisis wrought by the Depression was likely an important causal force.

More to the point, if the growth of the government was a continuation of a then-

ongoing trend, then the presumed importance of the Depression period would be

attenuated.

The restrictive labor laws arising from the boomerang against unions ensured

that corporate tyranny would prevail on the shop-floor and, necessarily, I will argue,

spill over into politics. The relationship between corporate power and worker resis-

tance is clear: the acquisition of profits necessitated continuous production6 and the

resultant labor quiescence contributed to the long period of annual growth that char-

acterized the the 1950s to mid-70s (cf. Lichtenstein 2012, ”Postwar Accord”).

In evaluating the labor relations policy domain, the unity and hostility of Amer-

ican owners is relatively high in comparative perspective (Vogel 1979; 1981). More-

over, scholars have tended to assume that this business unity has led to political

strength. But the American experience does not support the claim that business

unity overrides popular majorities; rather, business unity is an imperfect predictor of

legislative outcomes (Smith 2000). Mass opinion during the period does not directly

support the claim that the citizens opposed Taft-Hartley or the drive to curb militant

unionism, generally (see Chapter Three).

Existing scholarship on progressive laws in the states (Patterson 1969; Shefter

1983; Bridges 1997; Weir 2005) focus on how state level variation in capacity and

political logics amplify or diminish the e↵ects of progressive interventions. Distinc-

6As did the war e↵ort.

137

tively, this study aims to justify a generalized conclusion in the area of labor relations.

Though I would not argue that the state-level politics were governed by an invariant

logic, I argue that the actors in each state did perceive clear and contrasting policy

and political goals in an equivalent fashion.

What is distinctive about this investigation is the aggregating of the state-level

outcomes and the attention paid to specific subnational experiences in probing ex-

planations for the distinctive American pattern of labor relations. Nations are all

exceptional in their own way but the US case is preeminent because of its global

role; and, more importantly, given the opulence of the American society, its nig-

gardly policies are particularly striking. Moreover, state-level variation in labor laws

suggests that deterministic explanations the emphasize unique institutional configu-

rations cannot obtain. The states, sharing political rules and circumstances broadly,

vary in the labor relations regimes they adopt; this militates against exceptionalist

claims.

As the Democrats were motivated by an interest in leveraging the New Deal

labor relations institutions to sustain the growing industrial organization and immi-

grant voting bases, the Republican o�ceholders were motivated by the demands of

business organizations and, importantly, farmer’s groups. In the following, I exam-

ine the statistical finding above that backlash to union disruption was predictive of

restrictive law adoption.

The statutes we have focused upon thus far are those that hamper class-based

union protest as opposed to those that hamper union organization, financing, or

political activity. These laws, though, were often passed in the same legislative session

if not the package, so these phenomena are closely related, if distinct.

138

A portrait of American trade union decline is incomplete without a regional per-

spective. The declines in the South and Southwest are the product of labor laws that

restrained new organizing and curtailed the scope of militance for existing unions.

However, attention to the historical record reveals that the political, social, and eco-

nomic conditions that produce legal regimes hostile to unions prevailed in states where

unions achieved a range of outcomes, some of them fairly positive. This would suggest

that union power is at most a decisive variable in explaining the adoption of laws at

some times and not others.

Again revisiting Table 4.8 above, it is evident that the ownership of farms

distinguishes Wisconsin and Pennsylvania. However, consonant with the statistical

results from earlier in this chapter, the rural representation in the statehouses were

the main bloc in retrenching legislation. Because the farmer-labor possibility was

a threatening one to the status quo this chapter will focus on Wisconsin and treat

Pennsylvania as a site for opportunistic comparison.

New York

The progressive capture of the New York executive was a defining feature of this

case. Already strong unions were best positioned to capture the gains of New Deal

legislation; in the New York case, the national anti-Injunction Act provided support

to garment worker unionizing.

Most labor militance in New York State occurred in New York City. For the

period under investigation and the years just preceding it, New York CIty accounted

for 70 percent of man-days lost - and during the early Depression it was more than 90

percent (Annual Report of the Industrial Commissioner, 1931). The unions based in

New York City, particularly the Clothing Workers (ACWA) and the Ladies’ Garment

139

Workers (ILGWU) actively lobbied for collective bargaining legislation. Moreover

the NYS Federation of Labor tended to actively endorse and support Democratic

candidates, in contrast with Gompersian voluntarism. And, indeed, the NYSFL

joined Governor Lehman in his call for union security legislation in line with the

Norris-LaGuardia Act (New York Times 8/15/33; 8/23/33).

And the garment workers were a potent force, given that the industry was

mostly small firms on tight margins. In the wake of a tri-state area general dress strike

in August 1933, Bernstein (1970) notes, ”The union’s power was so complete that

employer opposition collapsed at once....The employers accepted the closed shop...”

(87). What is more, just a month previous, Secretary Perkins had to prevail upon

ILGWU leader David Dubinsky to postpone a strike when federal-level negotiations

over an industry code broke down (New York Times, 7/25/33).

The state of New York had a sharp urban/rural cleavage reflected in the legis-

lature. The industrial areas of the state were Democratic strongholds and the less-

industrial areas were held by the Republicans. The Democrats were strong in the

New York City and the surrounding areas as well as the cities throughout the Hud-

son Valley and in the western part of the state. The Republicans represented largely

rural districts and benefited from malapportionment. Thus, it was nearly impossible

for organized labor to capture the seats they did not already hold.

Not unimportantly, elections in New York were governed under fusion rules;

that is, a candidate could be endorsed by more than one party. Thus, the centrist

parties can be pressured by the radical candidacies without those campaigns or their

supporters being concerned that their vote would only serve to get the less desirable

centrist party elected. In New York, the American Labor Party played that role on

the left in New York politics.

140

After passing the state anti-Injunction Act in the previous session, Governor

Lehman said in his State of the State address that ”I stand ready to exercise my

veto power and my right to appeal directly to the people to combat any plan to

destroy what I sincerely believe to be the finest set of social legislation ever enacted

by any State in the country. The State of New York has now reached a happy

position” (quoted in New York Times 1/2/36). And later the same year, Governor

Lehman reiterated his support for pro-union legislation at the State Federation of

Labor Convention (New York Times 8/26/36). And in the next session, the State

Labor Relations Act was adopted.

The first enumerated unfair labor practice was that of interference, restraint, or

coercion of employees in their attempt to exercise collective bargaining rights. This

concept is a direct intervention against the owners most basic strategy in the face of

new organizing. The decision to whether and how apply penalties or issue judgments

on the basis of this unfair labor practice, because of its lack of precision, is a residue

of political contestation. In its first Report, the New York Labor Relations Board

summarized their concerns thusly

It is not possible, of course, to establish rules as to when a discharge willand when it will not be held a violation of the Act. Each case must restupon its own facts and upon the inferences which may logically be drawntherefrom. While it is possible to list the various considerations which havea↵ected the decision of the board in a particular case, no one of them alonecan ordinarily be regarded as the determining factor where the Boardhas made a finding of discriminatory discharge. Invariably, the evidencereveals the presence of a combination of circumstances which, taken inconjunction, may lead to the conclusion that the discharge involved in theproceeding was in fact discriminatory (1939, 41).

After the progressive electoral victories in 1936, the Doyle-Neustein bill - the

State Labor Relations Act (SLRA) - passed the Senate and Assembly overwhelmingly

141

(New York Times 5/5/37). And the labor board created by the bill would itself be

constitutive of the subsequent labor relations context. It should be noted that in

the case of any of the Labor Relations Boards, the first e↵ort was to find a way to

compliance without resort to hearing or charges. However, the new Board would, like

the national Board come to be seen as pro-CIO.

The SLRA was based o↵ Wagner. However, unlike Wagner, which listed five

unfair labor practices and left the interpretation to the triers of fact (read NLRB),

the SLRA enumerated ten unfair labor practices and defined them broadly. Also,

the SLRA contained a craft unit proviso (a majority of employees in a particular

craft could make the craft a bargaining unit) and the SLRA permitted employers to

petition for an election.

The NYSFL would not continue its support of the ’happy’ labor relations

paradigm erected in New York’s Little New Deal. In the first two years of the SLRB’s

existence, there were ten contests where the CIO and AFL were directly competing for

recognition and the Board found for the CIO in seven of the cases while the AFL won

just one (in one case neither won employee majority and in another an independent

union won recognition) (New York Times 9/18/39). As the NLRB came to be viewed

as an enemy by the AFL (Gross 1981, 246), the NYSFL would come to oppose the

SLRB. Indeed, after a crucial transit workers bargaining dispute was ruled against

the AFL, George Meany declared the SLRB was seeking to impose ’Nazi-Communist’

regulations upon labor (quoted in New York Times 12/7/39).

Where the state AFL a�liate was opposed to the operation of the State Labor

Relations Act, the state CIO had ’not one single complaint’ regarding the admin-

istration of the act (quoted in Kim 2000, 233). On August 28, 1937, the Triboro

Coach Company entered into a collective bargaining agreement despite the CIO hav-

142

ing collected sign-ups from more than 109 of the 181 workers there. The State Labor

Relations Board ordered an election which the CIO was favored to win. The AFL was

incensed, having seen the Board put aside a previously existing collective bargaining

agreement. And at its annual legislative conference in April 1939, the NYSFL con-

demned the Board’s resolution as being prejudicial against the AFL and beyond the

color of law. The NYSFL would, subsequently, lobby in support of amendments to

jurisdictional strike laws and make common cause with the anti-labor Republicans

led by Irving Ives, a Republican state senator and later a US Senator who would vote

for Taft-Hartley (Ibid 241-3).

The Ives Committee, a parallel to the Smith Committee, was formed after the

conservative surge in 1938. With the support of the NYSFL, the SLRA was amended

to weaken its union security provisions: crucially, the preamble was amended to pro-

mote ’partnership and cooperation between workers and employers;’ and established

arbitration agreements as having the force of law. This latter point is crucial, it

encumbers workers’ legal recourse to protest grievances. The upshot of the amend-

ments, as Dubofsky concludes, ”Like the Ives bills, which weakened the New York

State Labor Relations Act and therefore diminished the power of organized labor, the

Taft-Hartley Act attempted to strip organized labor of what they had gained though

the New Deal law and labor policy” (1994, 204).

Unlike the Smith Committee, the Ives Committee - formally, the Joint State

Legislative Committee on Industrial and Labor Relations - was convened by Repub-

licans, rather than by co-partisans of the administration. However, the di↵erences

between the Dixiecrats and FDR were not unlike the di↵erences between Governor

Lehman and the New York Republicans; it also should be noted that a substan-

tial number of assembly Republicans voted for the SLRA much as House Dixiecrats

143

supported Wagner (Red Book). Again distinctively, the Ives Committee was de-

signed to make ”recommendations designed to increase industrial peace” in contrast

to the anti-communist basis for the Smith Committee. In his Story of a Legislative

Committee, Ives writes that ”At the beginning of 1938, a↵airs between workers and

employers seemed to be getting out of hand. Bad administration of the Wagner Act

was threatening the proper administration of our SLRA which, with the backing of

both industry and labor in the state, had been enacted in 1937. So the Legislature

created a Committee of eight members for the purpose of finding out what could be

done by law to bring about a happier relationship between employers and employees”

(1944, 3-4).

Both Committees set out to rewrite the preambles of the protective labor statue

it assailed and move policy in a more pro-capital direction. Both committees were

interested in undermining the pro-labor ’bias’ of the Labor Relations Board. The Ives

Committee produced a number of bills, and unlike the Smith Committee’s recommen-

dations, some were adopted, including a new preamble and new employer guaranteed

rights to oppose unionization during elections (Ives 1944, 8-9). However, the provi-

sions that served to restrain militance - other than the ban on jurisdictional strikes -

were in fact opposed by the NYSFL.

However, even in the New York case where the American Labor Party (ALP), a

third party formed by the CIO’s John Lewis and Sidney Hillman (a New York garment

union leader), stemmed the conservative tide in 1938, the state LRA was still amended

in line with the Ives Committee recommendations. Ultimately, in the same way that

the AFL, Dixiecrats, and Republicans in the Congress formed a majority for the

adoption of hostile amendments to the Wagner Act, the NYS Federation of Labor and

Republicans formed a state-level majority to adopt employer-friendly amendments to

144

the SLRA in the form of the Ives committee recommendations, which is analogous to

the contemporaneous Smith Committee at the national level. Moreover, as the Taft-

Hartley Act does concede that organized labor would be a factor in labor relations,

the Ives bills likewise accepted the fact.

Taken together, the New York retrenchment was swift but blunted by Lehman’s

governorship and the American Labor Party in the Assembly. Recommendations to

alter rules by which the Labor Board adjusted disputes were chief among the Ives

Committee proposals that were adopted. The New York case stands in the middle

between no retrenchment at all and a full retrenching bill. While strike rights were not

substantially curtailed, there were conservative amendments. The urban core of the

state legislature was su�ciently numerous to resist legislative attempts to contract

labor’s capacity for action. These outcomes dovetail with the regression results from

the earlier in this chapter.

Strikes were not more popular in the Northeast, as we observed in Chapter

Three. But the electoral ramifications were blunted. Nor could the local newspapers

have been said to be neutral in their coverage of labor stoppages, as we observed in

Chapter Two. Referring to the appendix in that chapter, we observe numerous exam-

ples of the Times and Herald Tribune declaring the strikes to be forms of insurrection

with the articles typically preferring to use o�cial sources or ownership sources as

opposed to those of the strikers. In the coverage from major papers in the East drawn

from the searches in the second chapter, the Times and Herald Tribune were carriers

of net negative source-messages.7 Overall, the New Yorkers were not in a distinct

information environment from those in Wisconsin, Utah, or the nation generally. As

7Sadly, the newspapers from Wisconsin and Utah are not amenable to digital searches; thereforean analysis of source-message content is not possible.

145

discussed in the second and third chapters, citizens were expressing attitudes more on

the basis of shared information and shared processing rules than they were operating

with di↵erent apprehensions of the state of the world.

Wisconsin

The Wisconsin SLRA followed the Wagner Act in its definition of unfair labor

practices for employers but not employees. And, like Wagner, a Labor Relations

Board was created to certify unions and investigate unfair labor practices. Unlike

Wagner, the Wisconsin Act specifically prohibited blacklisting and spying. Moreover,

the Wisconsin Board had the mandate to prevent and settle labor disputes, unlike

the Congress, which gave the NLRB no arbitration powers.

As a result of the SLRB, like the NLRB, there was a surge in new organizing,

particularly on the part of the CIO. And CIO new organizing would split the Farmer-

Labor Progressive Federation as cooperatives run by farm groups in the coalition were

targeted for organizing (Holter 1990, 35). This split contributed to the demise of the

majority coalition that had enacted a bill more pro-union than Wagner and paved

the way for the Republicans to take the majority in both legislative chambers.

The 1939 amendments to the Wisconsin SLRA presaged Taft-Hartley in nu-

merous respects. The Board lost its power to determine bargaining units - that is,

the Board lost its power to decide what the bargaining unit should be. As a result of

the defanging of the Board, its caseload dropped precipitously - unions seeking me-

diation or redress for grievances dropped by more than 90 percent (Ibid, 43). Also,

it allowed for employers to call an election at any time. The amendments instituted

a 2/3rd majority requirement for the union shop. Another important provision that

antedates Taft-Hartley was the ban on the secondary boycott.

146

In Wisconsin, the SLRA was adopted by a coalition of farmers and progressives.

However, the intervening strikes after the passage of the law caused the farmers to

leave the coalition (Ibid, 37). However, just as CIO new organizing in the South

doomed the union security majorities in the Congress, CIO new organizing in trans-

port and dairy doomed the Farmer-Labor Progressive Federation. Nonetheless, the

farm interests were contingent in their support:

...exemption from the scope of this bill [Wisconsin Labor Relations Act]cannot guarantee freedom from the disastrous e↵ects, due to the perish-able nature of our products, that labor disputes would have upon agricul-tural cooperatives. It is evident that exemption does not spell immunity.In view of this, these farm cooperatives withdraw their demand for suchexemption. In doing this, we hope that labor will appreciate the fairness ofspirit on the part of agriculture and we hope that such mediation machin-ery may be provided which, in the event of trouble, will be su�ciently fair,e↵ective, and quick to meet the unusual problems that would inevitablyresult from a tie-up of our perishable farm commodities (Swanton to StateSen. Earl Leverich 3/31/37, WCAC Files, Box 8, Folder 1).

The Wisconsin Farmers Cooperative Confederation, along with the Wisconsin

Council of Agriculture Cooperatives, was the preeminent farmer’s organization in the

state in the 1930s. Milo Swanton, their Executive Secretary, said to to the state

Manufacturers’ Association in 1938,

Unlike the industry manufacturer, the agricultural manufacturer main-tains a constant production at a variable price. As an employer the farmeris also in a very di↵erent position because of his very seasonal productionand because of the perishable nature of his products. An industrial strikeis a loss to all concerned....However, the farmer with goods that perishin a few days or hours and with seasonal production is in a weaker posi-tion as a bargainer with labor than the industrial employer. It is for thisreason that the Wisconsin Council of Agriculture, seeing the advance oforganized labor into the field of agriculture and carrying with it industrialtrade union policies, asked for a revision of state laws.(Wisconsin StateHistorical Society, WCAC Files, Box 3, Folder 8)

147

What is important in this passage is the fact that the Wisconsin farmer groups

moved away from the union protection coalition into the restriction coalition because

of the vulnerabilities to industrial disputes that were realized under the (baby) Wagner

law. The exercise of power by workers in Wisconsin undermined their future capacity

to exercise that power because it hurt the forces who were indispensable to the passage

of protective laws. Swanton, in private correspondence to an editor argues that

The farmer is both a capitalist and a laborer; and while in the past he hasbeen inclined to be sympathetic with labor, he has also come to realizethat he has a large investment in land, machinery, livestock, and buildings.He is a purchaser of both consumer and capital goods. He has come torealize that labor policies and high rates have a definite e↵ect upon hiscost of production and that labor disturbances frequently invade his rightsand his opportunities to make financial progress (Ibid, Letter to Roy Park,February 6, 1942, Folder 10)

The previous two chapters proposed that there was a bulwark in mass opinion

against strikes that would increase workers’ relative power that coincided with pre-

vailing unfavorable coverage of strikes and strikers. Here, though, we can see that the

interests of would-be coalition partners with labor evidently diverged. The paths to

forestalling Taft-Hartley were occluded in two ways: the citizens qua consumers were

opposed to the costs of strikes; and, incumbent coalition partners were a↵ronted by

those selfsame costs.

Unlike New York, the labor militance was dispersed throughout the state. The

formation of the third party came from the radical farm groups that finally found

an ally in the progressive wing of the state Republican Party (Milwaukee Journal

3/4/34). The bill was passed by the Farmer-Labor majority after more than one

hundred roll calls (Backstrom 1956, 332). However, the result would not hold.

CIO success came at the cost of farmer groups who had tipped the majority of

148

the state legislature toward labor interests. The CIO endeavored to organize coop-

eratives run by farm groups. These farmers, aggrieved by the concomitant strikes,

realigned with the Republicans after the 1938 elections and then they introduced a bill

to promote ’employment peace.’ The Employment Peace Act changed the processes

of unit determination, reducing the scope of Board activity in this area, in the same

ways as the Taft-Hartley Act did for Wagner. Moreover, like Taft-Hartley, the Act

outlawed secondary boycotts. This meant that unions could not operate in concert

to disrupt production. Atomizing unions in this way kept them from operating on a

class basis. More subtly, the theme of CIO new organizing disrupting long-existing

patterns of economic and social domination leading to harsh retrenchment is a story

that occurs outside the American South.

The Roosevelt recession was also costly to farmers: ”Butter is the same in price

as twenty-five years ago, while the wages of skilled workmen are, roughly, fifty percent

higher. Can the farmer help to boost wages which he has to pay, while his own income

stagnates, or goes lower?” (Swanton address to Milwaukee Chamber of Commerce,

WCAC Files, Box 3). And, these losses exposed the rift: ”It is true in a country like

ours, with surplus farm production, that, when the farmers are prosperous the rest

of society is prosperous. This is true because a surplus of farm product cannot be

sold at good prices to unprosperous customers. The converse of this statement is not,

however, equally true.” (Ibid, ”Mutual Relations of Agriculture, Labor, and Industry,

WCAC Files).

The opportunity for a farmer-labor alliance, then, was very di�cult to exploit.

And, what is more, antipathy toward Lewis would make forming such a coalition

di�cult. In the Hoard’s Dairyman, a predominant industry newsletter in Wiscon-

sin and Minnesota, an editorial argued that ”control of food is a weapon. John L.

149

Lewis knows this. Any man who can control the distribution of food can control the

destinies of the people. The organization of farmers into unions controlled by labor

dictators would take away from farmers freedom of action, and could conceivably

put farmers in a strait jacket from which there would be no escape” (WCAC Files,

Hoard’s Dairyman, 1943)

In the way that Operation Dixie proved a grave threat to the Roosevelt New

Deal coalition, CIO farm organizing proved a grave threat to Governor LaFollette’s

little New Deal coalition of labor and farmers. And, indeed, progressives, who were

Republican defectors went back to the Republicans in the 1938 elections. As Thomas

Amlie, the progressive Congressman, pointed out:”Although organized workers did

not vote for Heil, neither did they enthusiastically work for LaFollette, having become

somewhat dissatisfied with the independent role the governor played in the matter of

appointments to various boards and commissions, and the failure on the part of the

governor to consult representatives of labor on plans of reorganization and matters

of public interest” (Thomas R. Amlie Papers, Memoranda, Box 2)

One might wonder about the awareness level of farmers, though. Curiously, a

Farmer’s Cooperative poll in 1944 found that more than half of the members were

unaware whether the co-op was required to pay taxes; only 26 percent correctly

answered that the co-ops were obligated to pay taxes (WCAC files).

The ascendant farmer-labor coalition elected with the support of organized labor

and the coattails of FDR’s reelection campaign, delivered a protective labor law. The

organizing that resulted from that law broke the coalition. The fractious relation

between agriculture groups and organized labor, particularly the CIO, ensured that

any progressive majority would need to be delicately ushered through the adoption

and administration of its legislative program.

150

The Midwest was not out of step with the rest of the nation in terms of opposi-

tion to strikes. The Milwaukee Journal has not been archived in a fashion conducive

to subject and keyword searching. However, as we saw in Chapter Three, the opinion

in the Midwest was in sync with the national disapproval of labor militance. The agri-

culture bloc proved decisive in the state legislature in a context where mass opinion

was cool toward labor militance. The crucial takeaway here, though, is the emergent

opposition between the farm coalition that was essential to the progressive law and

the CIO unions strengthened by that law.

Aside from the fracturing of the farmer-labor majority, the sit-down strikes in

urban Milwaukee were themselves highly disruptive. Over the course of the first year

of WLRA administration, there were 28 sit-down strikes in Milwaukee. Nine hundred

workers in the textile a�liate of the CIO sat down at the Rhea manufacturing plant

for 17 hours. The garment workers held the factory throughout the night and saw

their wage and workplace safety pleas answered the next day (Holter 2001, 214-5).

But, over the course of the year, the sit-downs became less a tactic for winning union

gains and more a tactic in furtherance of shopfloor protest (Ibid 215).

Utah

Though we have been focusing on the cases of protective laws being wholly or

partially retrenched, we now turn to a case where the protective statute remained

intact. Indeed, in contrast to New York and Wisconsin, a restrictive statute did not

even make it onto the legislative calendar following the 1938 elections (Hinton 1975,

160)

Utah was ravaged by the Depression, having the fourth highest unemployment

rate in the nation in 1934. Further, the mining industry, which was the basis of the

151

state’s economy was decimated by the Depression, seeing a decline in mineral value

from approximately 122 million dollars to 22 million (Ibid, 4). And, Importantly,

the Mormon First Presidency was in favor of capitalism and against socialism in

principle. And all Mormons are beholden to the directives of they First President

(this is a more serious commitment than say a Catholic has to the Pope). The Mormon

Church cannot be said to have dictated political behavior during the early statehood

years but at least some politicians did feel su�ciently beholden to change their votes

on spending bills and Senate elections - and in one case even changing parties and

switching again on church orders (McCormick and Silito 2011, 439-42).

These distinctions do not render Utah unsuitable for comparison with the other

states. Though it is a state with a religious power parallel to the civil government

(more on this in a moment), it is not as if the Mormon domination of state politics

had subsumed political conflict, as there was sti↵ partisan competition between the

Democrats and Republicans since it had acceded to the United States. Also, it is

important to note that the state was ravaged by the Depression and was a major

recipient of relief expenditures. Ranked ninth among the states in receipts from fed-

eral spending from 1933 to 1939 , Utah benefited from New Deal programs. This is

attributable to Governor George Dern’s having been appointed FDR’s second Secre-

tary of War, though Utah’s unemployment rate was more than twelve points above

the national average (Cannon and Embry 2009 115-6).

The Mormon hierarchy largely opposed the New Deal and sought to get people

o↵ the relief rolls and put them on to the Latter Day Saints Security Plan (Cannon

2004; Darowski 2009).8 On the other hand, some Mormons helped people enroll

8The LDS scheme is extant, rechristened as the Welfare Plan in 1939, though most of its clientsare located in the third world nowadays.

152

in FDRs relief programs at the outset to the New Deal. With respect to foreign

policy, the Mormons supported FDR in lockstep; favoring intervention in the second

world war and could be reasonably described as patriots. The long-time Senator

Reed Smoot (R-UT) was an apostle and was openly endorsed by the First President

leading observers to draw a link between the Mormon Church and the Republican

Party (Hinton 1975, 183). However, the failure of the Church to forestall prohibition

repeal as well as the delivery of Utah’s electoral votes to FDR in 1936 and 1940 go

to demonstrate that while the Church was deeply influential, it was not able to set

policy as it wished. Though the Mormon Church issued statements in opposition

to the seeming unconstitutionality of the New Deal, it did not seek to erase federal

contributions to the state budget (Ibid, 190). Broadly speaking, we can think of the

Mormon Church as an influential pressure group during this period, rather than as a

dictator.

That said, the Mormon Church o�cials tended to oppose the union shop pro-

visions that were commonly a part of protective labor laws. Not only were Church

o�cials typically businessmen, there also existed the fear that solidaristic unionism

would supplant the Mormon faith (Powell 2013, Chapter 9). Nonetheless, when Lewis

called on the United Mine Workers to organize in Utah in 1933, the question was not

whether the mines would be represented but whether they would be represented by

an independent union or the CIO. Ultimately, though citizens were wary of resultant

price increases, the New Deal Democratic majorities ensured that protective labor

laws would stand (Ibid). Indeed, the New Deal would lead to the election of Demo-

cratic majorities that would govern uninterrupted until 1970 (May 1987, 181). This

is due in large part to the benefits Utah received from New Deal spending: Utah

was ninth in per-capita spending among the states and the state’s own Chamber of

153

Commerce lobbied against decreased Civilian Conservation Corps federal spending in

1938 (Hinton 1986, 276-6).

While party control did not change from the Democrats until well after the

War, Utah would abandon its protective law regime after Taft-Hartley’s passage in

1947. Likely between the influx of Black soldiers and the national Republican turn

against labor, the Mormon leadership saw an opportunity to press against socialism

which was opposed by the Church hierarchy prior to the New Deal because ”Social-

ism endangered capital; God’s kingdom should be one where business interests and

investments were safe; Mormons should champion the capitalistic system and the

values on which it was based and work within the two-party system to support it”

(McCormick and Silito 2011, 421). Though, in all fairness, the First President op-

posed FDR’s reelection in 1936, after the church had enacted its Security Plan and

Roosevelt carried Utah by 70 percent of the vote in that election and would gain more

than sixty percent of the vote in the next two races (Frankoski 2009, 169).

Though the adoption of the Wagner Act impelled further organizing in Utah’s

copper and silver mines, the resultant chaos did not spill over into the major conurba-

tions. The general strike in Eureka, Park City, and Bingham mines from October 10

to December 16, 1936 resulted in property damage and violence between picketers and

strike-breakers and led the (Summit) county Sheri↵ to request that the Democratic

Governor Blood call in the national guard. In response, the governor sent a skeleton

force and mediated the dispute. The resultant wage increase was hailed as a victory

for labor (Deseret News 12/8/16, 1; Hinton 1975, 217). There was some organizing

among taxi drivers and auto mechanics in Salt Lake City, the scale of this organizing

was in no way comparable to the general strikes that crippled major metropolitan

areas in the East and Midwest.

154

Thus, the Utah strikes were not particularly disruptive because the radical

unions in the mining industry did not generate an interested constituency that would

leave the protective majority to help form a retrenchment coalition. In contrast to

the logic that prevailed in Wisconsin, there was no farmer interest to overcome and in

contrast to the logic that prevailed in the New York, there was little urban citizenry

to provoke with shortages, nor would workers have been inclined to impose such

shortages on people they considered brethren as more than two-thirds of the state

was Mormon in the 30s.

Importantly, the state was homogeneous in its population; the state was the

whitest in the nation in the 1930s. Interestingly, the state had an influx of black resi-

dents during the war as more than a million black GIs moved to Utah. This might be

a cause for the Utahan move from protection to restriction, which would, of course

be consistent with the received wisdom on New Deal retrenchment. But the homo-

geneity decreased the threat implied by radical worker militance because the Mormon

church was not concerned about being politically or economically marginalized. Put

another way, if there were some disruption, there was no reason for sweeping backlash

because the mining unions would never have been ascendant in Deseret.9 Though the

most widely-circulated newspaper in Utah o↵ered more contextualized coverage of

strikes than most other presses, the overall tenor of the coverage was negative in line

with the articles analyzed in the second chapter. Altogether, it is more likely that

Utah’s distinctive policy outcome was the result of a comparatively lower social cost

to strikes than it was of a distinct information environment.

The state of Utah stands as the only state that did not enact a retrenching law

9The name the original Mormon settlers gave to Utah. The Deseret News is the name of thenewspaper the Church has owned and printed continuously since 1850; it was and remains a goodsource for the o�cial line of the LDS chieftains.

155

after having adopted a protective law during the working class interlude. Though

state urbanism was previously found to be negatively associated with restrictive law

adoption, the Utah case is not particularly in tension with this finding because of

the relatively muted worker protest. This view is, however, in tension with the core-

periphery distinction drawn above. Utah was decidedly on the periphery, in contrast

to the industrial core states that adopted protective laws, yet, unlike the periphery

states in the Southwest, it maintained a protective labor relations framework until

the Taft-Hartley amendments. Overall, this case suggests that there was a qualita-

tive di↵erence in the e↵ects of militance depending on the arrangement of partisan

coalitions ex ante. The apparent uniqueness of the Utah case indicates how it is an

exception that proves the rule: strike retrenchment was avoided where the strikes

themselves were not disruptive.

Discussion

What is being addressed by the historical comparison is the relative impor-

tance of union activity in politics conditional on partisan composition. Students of

the laggard status of the American welfare state pay special attention to political

institutions, particularly majoritarian electoral rules, federalism, and separation of

powers. But a systematic comparison of the states challenges these explanations.

Even in contexts where labor was politically strong, it su↵ered setbacks. Also vari-

ant across cases - though increasing over the 1935-47 time frame - was the unity of

capitalists.

Taken together, in some sense the labor law outcome is a product of partisan-

ship. However, partisanship does not carry a uniform meaning. Regional di↵erences

imply di↵erent coalitional fractures and alignments. For example, the farmer-labor

divide (or coalition), was put under varying degrees of tension; whereas the in New

156

York State union density and a pro-union executive, itself elected by largely Catholic

majorities, combined to create a circumstance wherein labor could achieve union se-

curity laws and blunt (if only somewhat) capitalist-led e↵orts at retrenchment.

Alternately, explanation arising from theories of political composition can be

understood as the residue of social and economic forces within the political unit.

That is, the partisan labels might not matter as much as the balance of the factors

of production and the degree of racial or religious fracture. But, again, attending

to demographics and economic factors yield heterogeneous results. The farmer-labor

coalition for progressive policy was fractious to be sure. However, it is a question of

degrees. And, the Utah and New York experiences suggest that di↵erent factors of

production can yield similar statutory results. Though Wisconsin’s legislature had far

less party discipline than New York’s - indeed, the Wisconsin lower house regularly

had cross-party voting on the vote to organize the chamber and elect the speaker

(Backstrom 1956, 339) - the partisan composition of the two 1937 legislatures that

adopted SLRAs did not a↵ect the substantive contents directly.

So, then, why were the protective laws adopted in the first place if they would

be so easily undone? As Patterson (1967, 79) wrote,

”The reform impulse of the 1930s unlike that of the progressive era wasfederal in origin and limited in duration. [...] since many key New Deallaws a↵ecting states – social security, the Wagner Act, and fair laborstandards – were passed in or after 1935, and since the Supreme Courtdid not sustain them until 1937 or later, many states were reluctant toenact ”Little New Deals” before 1937. Then the conservative reaction of1937-38 descended and the main chance was gone.

Moreover, subnational Democratic parties were as fractious as the Party in

Congress (Ibid, 80). The factionalism within the Party was at times the outgrowth

of liberals confronting long-standing political machines and sometimes it was provin-

157

cial selfishness in the state chambers but the general upshot was to ensure that a

unified Party program would not be adopted.10 The initial adoption of a protective

labor law required a then-novel concept of the American state intervening in labor

organization.

In practice, the socio-political coalition for progressivism in Wisconsin had to

be a farmer-labor alliance. The majority of the political unit in most cases could not

be unionist, so alliance would be necessary. (This is the logic that labor relies upon

in lending support to the civil rights movements of the 1960s while subordinating the

crucial goal of 14(b) repeal). In order to build an alliance with the farmers, labor

had to o↵er inducements. From a programmatic standpoint, labor and agriculture

could agree on the necessity of aid from the public till during the Depression. But

the class actors needed a class vehicle. If a class is constituted by persons sharing

utility functions, then it is evident that labor laws applied to agriculture result in

a coalitional tension. That is, the direct result of labor militancy implies a loss for

agriculture insofar as transport and workers are involved.

In this analysis, the equivalency premise takes a di↵erent shape. Across political

units there is a struggle for internalization of the union to the Democratic Party or

the Democratic Party to the union in each case. The extent to which unions are

internalized, I expect, depends in part on the degree to which the AFL (dualist) can

resist the CIO (monist) unions. In Wisconsin, the monist unions needed to coalesce

with farmers to form a majority. That is, the rural urban cleavage is dominated by

New York City in New York while Madison and Milwaukee had no such preeminence

10Not to suggest that had FDR actively sought to purge uncooperative Party members at thestate level, the Democrats would have developed forty-eight New Deals. Though FDR was highlypopular after his first reelection, there would have been entrenched and resentful forces opposingsuch an e↵ort.

158

in popular terms by comparison.

What comes through clearly from considering the cases is that the CIO new

organizing was destabilizing. In each unit under consideration, CIO new organizing

and growing rivalry with the AFL exacerbated fault lines in the majority coalitions

that adopted union security legislation in the first place. The labor law became

constrained as a result of its success. This implies that the reason for the adoption

for restrictive laws is the upshot of the adoption of the protective laws. The character

of restrictive provisions adopted by the states of New York and Wisconsin also speak

to the coalitional costs of labor action: New York adopted a ban on jurisdictional

strikes while Wisconsin was sure to impose a cooling-o↵ period on agriculture-related

labor stoppages.

The media and public opinion probes in Chapters Two and Three revealed

that the prospect for worker protest finding public support was minimal; and, fur-

ther, that the media portrayal of strikes only served to limit that prospect. This

chapter advanced the overarching argument of the dissertation in that it shows why

the unions could not have forestalled Taft-Hartley under any plausible circumstance.

The various states suggest strongly that the pathway to a workplace rights regime like

the ones that govern the a✏uent counterparts of the US was a foreclosed possibility,

even at the height of industrial union strength. Though the states provided varying

political opportunities and threats, the result was the same across varying contexts.

The unions could only avoid retrenchment in a political context wherein workplace

resistance was not especially costly to preexisting political actors.

Attention to these states indicates that the intimate relationship between the

Democrats and the unions was not the result of the union leadership having been

co-opted or their having missed the potential of an alternate partisan coalition with

159

farmers. Furthermore, the successful resistance to the restrictive amendments in New

York in contrast to neighboring Pennsylvania suggests that the rules of the game

determined labor’s position in partisan competition. The shrewd progressive, Amlie,

explained why voluntarism was a non-starter by the mid-40s,

The day is over when labor can a↵ord to take the position that in Congres-sional contests it will support its friends and oppose its enemies. Unlesslabor takes the responsibility of brining its friends into the political arenato run for o�ce by undertaking to carry the financial load involved, laborwill have no friends to supports. This has been strikingly brought home inthe present [1944] campaign. According to Governor Benson of Minnesota,there are only two candidates for Congress that liberals can properly sup-port in Minnesota this year. In the other seven Congressional districts,the candidates who won the Democrat-Farmer-Labor nomination are nobetter than their Republican opponents. In my opinion, the situation isalmost as bad in the other mid-west and northwestern states.”(ThomasR. Amlie, ”Memo to Sidney Hillman,” ibid)

It appears more likely that Hillman and Murray made a strategic calculation

that was basically correct: failing to wed with FDR and liberal Democrats would have

likely ensured that labor would lose all its Depression gains in the post-war period,

much as it did after the First World War (Gall 1989, 24-6).

The new institutions that enshrined collective bargaining in the law promoted

disruptive militance which generated an electoral backlash. These labor boards al-

tered the preferences of industrial unions because accepting the benefits of legal pro-

tection implied submission to subsequent amendments. To order large-scale extra-

legal militance like the organizing drives of the CIO during its break from the AFL

after Wagner being ratified by the Court would have cost the CIO its relationship

with the president they paid dearly to elect. And the very acquiescence of the Court

to the packing scheme itself further suggested that the CIO was well-advised to main-

tain its close alliance with the Roosevelt administration. Congressional conservatism

160

within the southern delegation combined with Republican hostility in the North also

demanded that the labor movement eschew voluntarism and work to alter the group

composition of the Democratic coalition.

But, just as importantly, Lewis’ voluntarism ran aground in its e↵ort to form an

alliance with farm owners. Due to their self conception as capitalists; their historical

primacy over labor; and, inclination toward patriotism for the war e↵ort, the farmers

in the Midwest were themselves unwilling to forge such an alliance. Though these

alliances were intermittently in power at the state level, the newly ascendant unions

of the 1930s would not have been able to form these coalitions because the militance

that empowered the new unions was unacceptable to agriculture.

The fact that heterogeneous states arrived at similar outcomes amidst a variety

of union postures suggests that alternate political paths were occluded. Organized

labor as a political vehicle was presented with four strategies vis a vis the parties and

polities. First, quietism; second, voluntarism; third, form the parties-in-chambers

with unionists themselves; fourth, become constituent members of the Democratic

Party, like the British Trades Union Congress and the Labour Party. Each of these

alternatives is freighted with costly tradeo↵s, and moreover, each was tried with

varying degrees of success. In the several states, the unions tried to sta↵ the parties

(MI), become central to the Democrats (PA, NY); play both ends against the middle

(WI, OR).

The Michigan unionists were riven by racial division and therefore were unreli-

able a a progressive force. Moreover, though the UAW became a predominant force

in Michigan politics, they were unable to unseat Republican Senator Vandenberg,

who opposed Wagner and helped override Truman’s veto until after the passage of

Taft-Hartley (Greenstone 1969, 111-4). In Oregon, the unionists were able to work

161

with Republican Senator Wayne Morse - who voted to uphold Truman’s veto - but

this course was freighted with di�culty. As Morse wrote to the CIO political director

Jack Kroll late in 1949, ”You people in the CIO are going to have to make up your

minds in respect to my candidacy for reelection as to whether or not you want any

liberals in the Republican Party. The attitude of you CIO critics in Oregon leads me

to believe you want liberals only in the Democratic Party” (quoted in Foster 1975,

197). It eventuated that the CIO had to work to only have liberals in the Democratic

Party because the third party option was not viable and because the Republican

Party had senior o�cials as hostile to labor as the Southern Democrats.

But standing firmly with the Democratic Party, even outside the South, did

not imply that the CIO could have forestalled retrenchment. In the states without

a segmented work force, the AFL held great sway and their opposition to state La-

bor Relations Boards and to the new organizing helped galvanize opposition to the

Democrats, particularly Governor George Earle (D-PA) who was unseated in 1938

(Davin 2000). And, the presence of fusion balloting in New York State enabled

Hillman’s American Labor Party to operate as an adjunct of the Democratic Party

without being subsumed by it and without siphoning votes o↵ it. That is to say, the

case of New York being without a significant retrenchment law (i.e. not one that in-

hibits most strikes) is not necessarily an implication that these comparisons indicate

a way forward for labor. The cases are more useful to rule out courses of action.

Nonetheless, to counsel pre-War labor to take the militant tack that John L.

Lewis recommended would have likely resulted in further legal repression, rather than

further programmatic gains. Taken together, the results tend to strongly indicate that

industrial labor would not have been better served by choosing a path of militance in

line with Lewis’ persuasion. This implies that labor cannot expect to form a majority

162

for socioeconomic reform via militance without forming alliances with other groups,

as standing on principle alone would more likely than not returned unions to the late

1890s. But this is something of a straw man, as Lewis did not counsel militance

alone but sought alliances with farm groups (Zieger 1988, 118-24; James A. Wechsler

Papers Box 1, WSHS).

From a strategic perspective, the unionists (particularly, the CIO) were not in

a dissimilar position with respect to the Democratic Party than they were vis-a-vis

capital. The withholding of the cartel’s services constituted the gravest threat the

union could muster. And, like in a strike, the workers’ resistance is highly costly.

What the unionists wanted to achieve in order to adopt social democratic policy was

to ensure that a majority of the majority coalition in the chambers were pro-union

voters. Because of the Dixiecrats, this was nigh-impossible in the Congress. But, in

principle, this goal was achievable in the statehouses.

Students of American labor politics have not failed to notice the crucial im-

portance of the CIO’s split from the AFL (Galenson 1960; Davis 1986; Rogers 1990;

Weir 2013) which divided the American labor movement into an industrial wing with

a Democratic partisan a�liation and a ’bread-and-butter’ wing that often acted at

cross-purposes with the other. What is distinctive about this dissertation is the sys-

tematic examination of the inability of the CIO to forge protective coalitions after

the schism. The split between the AFL and CIO came to define labor politics from

1936 onward. Indeed, the fragmentation of American labor is a major reason the

United States has never known a viable socialist party. As Zolberg (1986) argues, the

American labor movement is the most decentralized and divided among the a✏uent

nations. But all labor movements are distinctive for they are embedded in distinctive

contexts (400).

163

The battle between the rival federations necessitated the use of jurisdictional

strikes, which certainly helped foster the impression that labor stoppages were as

much a tool for racketeering as they were a tool for economic or social change, writ

broadly. Indeed, this line of argument featured in the Senate debate in favor of the

ban on secondary boycotts. However, in areas where the transport union (Teamsters)

engaged in secondary boycotts and hot cargo boycotts, these strikes were more often

mafioso tactics than class-organizational.

Moreover, the fragmentation of American unions is a mirror for the contestation

between class sympathies and particularistic benefit-seeking. The CIO was more

designed as a class vehicle but its growth was dependent on the occupation of o�ce

by liberal Democrats. As a result, the CIO became internalized to the Democratic

Party. The AFL, being threatened with marginalization, necessarily had to pose

itself in opposition to the nascent labor-left coalition. Without a unified, or at least

undivided, labor movement in politics, there did not exist a feasible majority for

union security measures.

Conclusion

The rapid failure of protective laws outside the South implies both that racial

segmentation is not what doomed the labor movement to adjutant status and that

militant unionism would have failed even if the US had not entered the War. The

cases in this chapter take a representative cross-section of the varying results that

labor enjoyed at the state level; where New York serves as an exception that proves

the rule. Fusion balloting in New York elections was and remains a crucial tool for a

third-party vehicle on the left.

An expanded focus on labor relations in the states takes the unevenness of New

164

Deal interventions seriously, which suggests that study of the New Deal order should

consider the national initiatives as being interwoven with di↵ering political units. In

a sense, the disconnections between the national AFL and CIO and political parties

and the state-level organizations and parties generated an ’intercurrent’ (Orren and

Skowronek 2004) labor relations order.

The overall role of labor law has been, on the one hand, to provide a platform

for asserting labor’s right to contend on an equal footing with capital but, on the

other hand, to limit the very capacity to engage in militant confrontation. The price

of legitimate access to the political arena is tactical moderation. But within this law

framework, labor and capital have structural dissimilarities that favor the owners.

Claus O↵e summarized the problem thusly,

For the sake of their power, unions are forced to maintain a precariousbalance between mobilization of resources and mobilization of activity,between size and collective identity, between bureaucracy and internaldemocracy. None of these dilemmas applies with comparable seriousnessto business and employers organizations for the reason that they do notdepend on internal democracy, collective identity, or the willingness toengage in solidary action, because of the very fact that they are alreadyin a structural power position which renders complications such as theseavoidable (O↵e 1985, 187-88).

In a sense, Plotke’s (1996) argument is a restatement of O↵e’s point in that

observers should be impressed with the achievements of the poorly institutionalized

labor movement in the midst of a relatively low-capacity American State. This point

is well-taken as the CIO (and AFL) was in a di�cult spot. It would not be wrong

to say that unionists faced a Hobson’s choice between allying with the Democratic

Party and increased militancy.

The only jurisdiction where the protective law was left intact was Utah, which,

as we have seen, was unique in that the newly empowered unions did not threaten

165

to undo the preexisting political equilibrium. That is, although the New Deal was a

challenge to LDS supremacy in broad strokes, the worker protest element in itself did

not constitute a grave threat as it did to the order in political units in the industrial

core, or racially segmented units in the periphery. The experience of the Wisconsin

case clearly illustrates the immense di�culty of constructing a farmer-labor coalition

after the New Deal.

But none of the above is meant to say that the Taft-Hartley Act would not have

been passed without the states having passed laws first. Statehouses were not full of

’hayseeds’ (Teaford 2002, 5) but they were not leaders in national policy on important

substantive matters. The reason states passed laws that anticipated Taft-Hartley

was, I have argued, because of the disruption to commerce and society wrought by

industrial strikes that politically marginalized radical organized labor.

166

Table 4.1: Strike and CIO Membership TrendsSources: Bureau of Labor Statistics and Troy and Sheflin (1965)

Year Strikers (1000s) CIO Membership (1000s)1928 314 –1929 288 –1930 183 –1931 342 –1932 3241933 1,168 –1934 1,467 –1935 1,117 3881936 788 1,0361937 1,801 1,9911938 688 1,9571939 1,171 1,8371940 577 2,1541941 2,362 2,6531942 840 2,4921943 1,981 3,3031944 2,116 3,9371945 2,615 3,9271946 4,611 3,9471947 1,629 4,4501948 1,435 4,4501949 2,537 4,314

167

Table 4.2: Penalized Maximum Likelihood Odds-Ratios ofRestrictive Law Adoption, 1936-1947, All 48 States

Variable Odds-RatioStrike Volume 1.001**

95% Conf. Interval 1.00 - 1.01GOP Veto 4.18**

95% Conf. Interval 1.68 - 10.41South 1.16

95% Conf. Interval .394 - 3.38Urban .972

95% Conf. Interval .941 - 1.00Black 1.01

95% Conf. Interval .982 - 1.05Income 1.85

95% Conf. Interval 1.19 - 2.86Immigrant .98

95% Conf. Interval .899 - 1.07N 576

Log-Likelihood 148.840Model p > �

2 .0064*p < .05 **p < .01 ***p < .001

Table 4.3: Penalized Maximum Likelihood Odds-Ratios ofRestrictive Law Adoption, 1936-1947, Non-South

Variable Odds-RatioStrike Volume 1.001**

95% Conf. Interval 1.00 - 1.002GOP Veto 12.55*

95% Conf. Interval 1.01 - 155.66Urban .956*

95% Conf. Interval .921 - .993Black .865

95% Conf. Interval .629 - 1.19Income 2.45**

95% Conf. Interval 1.31 - 4.58N 361

Log-Likelihood -83.528Model p > �

2 .0056*p < .05 **p < .01 ***p < .001

168

Table 4.4: Marginal E↵ects on Restrictive Law Adoption, United States, 1936-1947

Variable dy/dx Z X

Strike Volume .2971* 2.34 8.96GOP Veto 1.32* 2.89 .578South -.031 -.06 .373Black .013 .82 9.33

Immigrant -.006 -.14 8.09Urban -.037* -2.14 46.8

Log-Income .224 .81 11.07Owner Occupancy .149 .08 .487

Farm Own .139 .95 9.14

Table 4.5: Penalized Maximum Likelihood Chow Test, Non-South DV=RestrictiveLaw Adoption

Variable (1)Strike Volume 1.55

95% Conf. Interval .999 - 2.41GOP Veto 16.29

95% Conf. Interval .922 - 287.78Break 1024.634*

95% Conf. Interval 1 - 2217Break*Strike Volume .69495% Conf. Interval .435 - 1.1Break*GOP Veto .06195% Conf. Interval .003 - 1.15

N 576Model p > �

2 .032Chow p > �

2 .019*p < .05 **p < .01 ***p < .001

169

Table 4.6: Odds Ratios, Pre- and Post-1939

Variable Pre-1939 Post-1939Strike Volume 1.008* 1.011GOP Veto 3.762* 51.921South 1.343 1.214Black 1.023 .977

Immigrant .983 .971Urban .976 .970

Log-Income 1.485 2.051N 192 384

*p < .05 **p < .01 ***p < .001

Table 4.7: OLS Estimates of E↵ects of Republican Seat Gainin 1938 State Elections

Variable (1) (2) (3) (4) (5) (6)CIO Density 58.79* 57.64* 60.2* 59.07* 62.35* 63.5*(Std. Error) (24.03) (24) (23.95) (24.65) .(24.42) (24.98)FDR Vote - -.456 -.504 -.496 -.211 -.222(Std. Error) (.43) (.429) (.438) (.479) (.495)

Urban - - -.174 -.21 -.224 -.194(Std. Error) (.150) (.191) (.188) (.282)

Black - - - .829 .080 -.056(Std. Error) (2.59) (2.6) (2.82)

Income - - - - -24.59 -24.48(Std. Error) (18.01) (18.42)Immigrant - - - - - -.112(Std. Error) (.766)

N 29 29 29 29 29 29Constant 13.55 40.63 52.04 52.37 128.83 129.23Adj. R2 .1512 .1550 .1658 .1347 .1648 .1277

*p < .05 **p < .01 ***p < .001

170

Tab

le4.8:

Sam

ple

Restrictive

Law

s

State

Year

Measure

Law

Tennessee

1937

Sit-dow

nBan

TennesseeCod

e1937,Title

160

Vermon

t1937

Sit-dow

nBan

Vermon

tLaw

s1937,Number

210

Pennsylvan

ia1939

’Bab

yWagner’Amendment

Lab

orRelationsAct

of1937,am

ended

Michigan

1939

Anti-Violence

Provision

Lab

orMediation

Act

Minnesota

1939

Sit-dow

nBan

MinnesotaLaw

s1939,Title

377

Oregon

1939

Jurisdiction

alStrikeBan

OregonLaw

s1939,Title

2,Section

1W

isconsin

1939

’Bab

yWagner’Amendment

WisconsinLaw

s1939,Title

57,Section

111

New

York

1940

Arbitration

StrikeBan

New

YorkLaw

s1940,Title

851

California

1941

Arbitration

StrikeBan

California

Law

s1941,Title

1188

Texas

1941

Anti-PicketingLaw

Texas

Law

s1941,Title

100

Marylan

d1941

Sit-dow

nBan

Marylan

dLaw

s1941,Number

340

Mississippi

1942

Anti-PicketingLaw

MississippiCod

e,Number

323

Alabam

a1943

Anti-PicketingLaw

Bradford

Act,Section

9Arkan

sas

1943

Anti-PicketingLaw

Arkan

sasLaw

s1943,Number

183

Colorad

o1943

Secon

daryBoycott

Ban

Lab

orPeace

Act

of1943

Florida

1943

Right-to-WorkLaw

FloridaCon

stitution

,Amen

dment5

Kan

sas

1943

’Hot

Cargo’StrikeBan

UnionRegulation

Act

of1943

Sou

thDakota

1945

’Hot

Cargo’StrikeBan

Sou

thDakotaLaw

s,Title

86,Section

4Arizona

1946

Right-to-WorkLaw

ArizonaCon

stitution

,Article

25Lou

isiana

1946

Anti-Strike

Lou

isianaActs1946,Number

180

Nebraska

1946

Right-to-WorkLaw

NebraskaCon

stitution

,Article

15Virginia

1946

Anti-PicketingLaw

Virginia

Law

s1946,Title

229,

Section

1Delaw

are

1947

’LittleTaft-Hartley’

Delaw

areLaw

s1947,Title

212

Georgia

1947

Right-to-WorkLaw

Georgia

Law

s1947,Title

34Idah

o1947

Secon

daryBoycott

Ban

Idah

oLaw

s1947,Title

206

Iowa

1947

Secon

daryBoycott

Ban

IowaLaw

s1947,Number

111

Maine

1947

Anti-ClosedShop

Law

MaineLaw

s1947,Title

395

Missouri

1947

Anti-StrikeLaw

Missouri

Law

s1947,Number

79North

Dakota

1947

’Hot

Cargo

Ban

North

DakotaLaw

s,Title

115

Ohio

1947

PublicEmployeeStrikeBan

Ohio

1947

Law

s,Title

261

Utah

1947

Anti-StrikeLaw

UtahLaw

s1937,Title

55,am

ended

171

Tab

le4.9:

Variab

lesan

dSou

rces

Variab

leTim

eVariant

Interpolated

Sou

rce

Protective

Statu

teYes

No

Killin

gsworth

1948

Restrictive

Statu

teYes

No

Killin

gsworth

1948

Republican

Veto

Point

Yes

No

Con

gressional

Quarterly

1994,667-716

Blacks

Yes

Mostly

Census

Urban

izationYes

Mostly

Census

Union

Density

Yes

Mostly

Troy

andSheflin

Union

Militan

ceYes

Mostly

BLS(W

awro

andKatzn

elson)

Farm

Owned

Lan

dYes

Mostly

Census

Owner

Occu

pied

Units

Yes

Mostly

Census

Sou

thNo

No

Katzn

elsonan

dcollab

orators

FDR

Vote

Yes

No

Presid

ency

Project

172

Table 4.10: Variable Summary

Variable Min Max Mean Std. Dev.Strike Volume .009 4223.71 53.24 200.05GOP Veto 0 1 .578 .494South 0 1 .37 .48Urban 16.6 92.4 46.8 18.64Black .1 50.2 9.33 13.06

Immigrant .4 26.3 8.09 6.5Farm Owned 5.21 11.39 9.14 1.13

Owner Occupied .29 .64 .487 .07

Table 4.11: Summary of the Cases

Wisc. New York Penn. Utah Mass.Protective Law 1937 1937 1937 1937 1937Restrictive Law 1939 (1940) 1939 1947 (1941)Change to GOP Yes No Yes No No

Strike Volume Rank 8 6 4 41 23Farm Owner Rank 14 27 28 13 15

173

Chapter 5

Conclusion

Putting It All Together

The previous chapters have made the argument that the retrenchment against

worker protest was a result in line with majority preferences and, indeed, was a fairly

popular result across the nation. Yet, this dissertation has been premised on the claim

that the abrogation of the right to strike was a depredation borne by the workers of the

United States. So, then, the democratic competence of the mass public is called into

question - but not from the public side; rather, from the elite side - because the mass

public can only be judged based on the information available to it. The newspapers,

though factual in their reporting, did prefer to emphasize the supply shortages and

public disorder arising from strikes than the grievances or class interests being upheld

by the labor stoppages.

In the second chapter, I turned to the information that citizens relied upon. I

examined the coverage of salient and consequential strike episodes during the New

Deal period as well as editorials and syndicated columns alongside automated keyword

174

searches. I established that militant union leadership received the preponderance of

coverage. And with respect to the content of news coverage, unfavorable items against

strikes were generally prevalent from the Flint strikes onward. Also, o�cial sources

became more denunciatory over the course of the 1930s and 40s. That the content

was analyzed episodically as opposed to continuously imposes a limitation on the

conclusions drawn. Though it is unlikely, there is a chance that the content of media

reports were more evenhanded than these samples suggest. More subtly, it is not

evident that, to the extent coverage was against unions, the news was leading or

lagging mass opinion. Whether editors and owners were ’chasing eyeballs’ remains an

open question. But I proceed with an emphasis on the finding that the coverage was

mostly unfavorable to unions when it plausibly could have been more balanced.

The third chapter dealt with surveys of the citizens that returned retrenchment

majorities to the statehouses in 1938 and the Congress in 1946. I examined surveys

regarding support for unions, labor militance, the Wagner Act and militant union

leadership and compared them on a regional and class basis. Although this disser-

tation has been somewhat agnostic on the puzzle of why the militance that led to

protective laws also led to restrictive laws, the third chapter regarding public opin-

ion data suggests a tentative answer: namely, that the public approved of strikes

in furtherance of bargaining unit organization but deplored strikes in furtherance of

working-class interest or in contravention of war aims. Though the opinion and media

data are only analyzed graphically, it is evident that there was a connection between

sentiment toward unions and volume of coverage. The survey data presented were

arranged as rather limited time-series datasets because of the paucity of repeated

survey items. Moreover, that the data were not broken out by state in some polls

- instead by region - implies potential conceptual slippage between the opinion and

175

media data on a sub-national basis. The sampling procedure employed makes impu-

tation of state level opinion di�cult. Nevertheless, we saw that the citizens were not

meaningfully divided along regional or class lines.

In the fourth chapter, we learned that the restrictive laws against unions were

adopted in order to delegitimize worker protest. Here I regressed state-level char-

acteristics alongside strike volume on the adoption of restrictive strike laws across

the nation and in the non-Southern states. The statistical results suggested that the

pattern of labor restriction was in some sense orthogonal to race, for the non-Jim

Crow jurisdictions also rapidly retrenched against unions. Further, the statistical re-

sults in that chapter suggested that the turning point in labor law history was in the

1938 elections. As noted in that chapter, the baby Wagner laws might not have been

adopted without the adoption of the Wagner Act in the first place, but the retrench-

ing laws did not need federal leadership to be adopted. Moreover, this chapter does

dampen potential observed variation in the laws in its focus on anti-strike provisions.

Union activity went beyond strikes, of course. What is gained in recompense is a

clarity of focus on worker resistance.

Alongside the work in these chapters, I compared states that elected to adopt

’baby’ Wagner Acts, asking whether and how the logic of the chapter applied to the

cases, and, in the fourth chapter, I asked why did they retrench so quickly. These

comparisons dovetailed neatly with the statistical results, showing that organized

labor coexisted uneasily with other members of the protective coalition1 while indi-

cating that the skilled workers contributed to the programmatic defeat of industrial

1Though it might be noted that farmer-labor coalitions had existed in both parties; the discus-sion in that chapter should not be taken as assuming that farmer-labor coalitions were inherentlyprogressive, though it was undeniably progressive in the majorities they it enjoyed in Wisconsin andMinnesota.

176

unionism. Moreover, the similar outcomes among heterogeneous industrial states is

in keeping with the key lesson of the mass opinion data: across regions, the public

was hostile to labor militance.

These findings imply that the anti-strike character of US labor law would have

found a genealogy outside the South, even if the Jim Crow delegation had not been

opposed to the CIO. The subsequent experience of the US labor movement demon-

strated that capital would grow increasingly intransigent as the usage and e�cacy of

the strike weapon waned. Strikes were not circumscribed to wage and benefit dis-

putes under a contract because other strikes are just o↵ensive to capital; they were

circumscribed because other strikes are o↵ensive to many in the mass public.

If there was in fact no connection between the unfavorable coverage that salient

strikes received and the unfavorable opinion toward militance that was broadly ex-

pressed, then we would have encountered a notable coincidence. The connection

between the unfavorable public opinion toward strikes and the adoption of retrench-

ing policies is perhaps more tenuous. However, I did find that strikes in Northern

states were positively associated with Republican seat gains in the 1938 elections.

And, given that the capitalist class was united in its opposition to labor, the absence

of broad public support for militance was a su�cient condition for retrenchment to

take hold.

Nonetheless, if these linkages did not hold, the core observation that the behav-

ior of the state legislatures in the 1930s took the character of a conservative reaction

to liberal strike policy furthered by the Wagner framework would also remain valid.

Furthermore, the core finding that this conservative reaction enjoyed broad tolerance,

if not favor, by the mass public would remain valid. The argument that the glass was

’half-full,’ also would remain valid, in light of the state cases.

177

As Lichtenstein points out, drawing from Daniel Bell, the Taft-Hartley law

amounted to a statement of power more than anything else (Lichtenstein 2002, 137).

Treating the restrictive legislative outcomes and the popular support they enjoyed

without reference to their normative content would be an error. The extant legal

framework that emerged from the New Deal is characterized by workplaces wherein

workers are subject to the arbitrary dominion of owners and managers. And, when I

say ’arbitrary,’ I say this because the workers do not enjoy any meaningful check on

the authority of management. The prevailing legal view is that workers are meant

to obey the dictates of owners and managers. This perspective is rooted in the com-

modification of labor and the common law traditions concerning disposal of property

(Gourevitch 2016, 315-7). In this chapter, I sketch an understanding of strike rights

as being inherently bound in the bundle of rights that make citizenship real. But,

one could just as easily follow Gourevitch in pointing out that the right to strike ”is

a right to refuse to do work while maintaining a right to the job. This conceptual

structure makes sense if we see the strike as a way of reversing the structural domina-

tion of workers as the most immediate, concrete point at which they experience that

domination: the threat of losing, or never acquiring, a job” (2016, 318). That is, the

right to strike is the right for a collective of workers to resist capitalist domination in

itself.

In the remainder of this chapter, I will turn to the question of the relationship

between strikes and freedom. The foregoing investigation of the delegitimizing and

truncation of the right to strike was not motivated by a perspective that is neutral

with respect to workers; nor was its central motivation an interest in the fortunes of

labor unions as such; rather, it was motivated by the idea that abrupt and collective

cessation of employment is an essential tool for securing the rights of workers, and,

178

by, extension, their political power. The phenomena of strike restriction is meaningful

because labor is central to American constitutional development (cf. Orren 1995) as

well as its political development, and is constitutive of much of civil society as well as

economic production. In other words, the fact of strike restriction is also value-laden

(cf. Putnam 2002, ’Chapter 1).

Strikes and Freedom, Reconsidered

The previous chapters have been preoccupied with delimitations on the right

to strike. In this chapter, I will be guided by two questions: What is the relation-

ship between strikes and freedom; and, how does the freedom to strike coexist with

prevalent concepts of liberty? This chapter will be a good deal more normatively

theoretical than the previous ones.2 We will be further interested in how the lessons

of the crucial New Deal period bear on the present political economy whose labor

market contains a large share of ’algorithmic’ jobs (cf. Collier 2015), and, therefore,

are patterned far di↵erently on the shopfloor than in the New Deal.

Therefore, given the information that citizens had, was their opposition to

strikes a wise attitude to express? In the present context, what is meant by ’wise’ is

a sound policy that coheres with democratic values. But in this instance, the policy

that coheres with democratic values is one that is incoherent with liberty. Drawing

upon David Hackett Fischer (2005, 4-10), observing the theoretical tension between

the right to be a free person fully as part of a society of free people (i.e. freedom) and

the right to be separate and independent (i.e. liberty). Further, English is the only

Western language containing this bifurcation (Ibid, 11). Opponents, particularly con-

servative ones, of protective laws that shield labor militance typically emphasize(d)

2Though I would suggest that it is impossible for a study of labor politics - or, truly, politicsgenerally - to be normatively innocent. For this reason, I reject the typical reticence of social scienceto eschew value-neutrality

179

the violation of the liberty of contract borne by the worker excluded by union. In-

deed, the problem with union dues was laid bare by opponents who argued that

paying tribute to any organization one disagrees with as a condition of employment

was an injustice. And this does not even contend with the urban scourge of union

racketeering. Witwer (2005, 528) argues that this narrative played a central role in

the late 1930s backlash against unions.

The investigation of the newspaper coverage of crucial labor episodes during

the New Deal suggested that the newspapers were likely to rely on o�cial sources

and tended to give less voice to labor sources. Moreover, the coverage featured de-

nunciatory statements against unions that usually were left unrebutted (examples in

the Appendix of the second chapter). So, then, the pessimism implicit in my pre-

vious work was on the bases of political knowledge and perception (cf. Shapiro and

Bloch-Elkon 2008): knowledge in the sense that the public was led to view strikes

from the perspectives of those most interested in maintaining labor peace; perception

in the sense that the strikes were framed as injurious to the war e↵ort, or a threat to

Lockean values of property and liberty.

Let us start with perceptual bias. Because strikes were framed by the press as

a direct assault on the American tradition (cf. Hartz 1955) and unpatriotic3 - leaving

aside the inconvenience wrought by strikes - the framing of strikes were commonly

depicted as having placed the US in the ’domain of losses’ (cf. Kahneman and Tversky

1973). That is, there was a mass cognitive bias toward viewing the costs of strikes,

rather than imagining the kind of society that would result from labor denying capital

the whip hand in socio-economic relations.

3This is not to say that the conservative tradition in American political thought is more ’essential’than the progressive tradition.

180

Dahl argued that ”if democracy is justified in governing the state, then it must

also be justified in governing economic enterprises; and to say that it is not justified

in governing economic enterprises is to imply that it is not justified in governing the

state” (quoted in Frymer 2010, 609). I will not attempt to evaluate this syllogism

here; however, I will argue that the extant imbalance in political power between

elite citizens4 and the mass public - so great that Page and Winters (2009) term it

’oligarchy’ - is a result of a demobilized and insecure workforce. This decline in worker

power was attended by a decrease in union membership and militance. Contrary

to prominent approaches in political science (Goldfield 1988; Getman 2010; Frymer

2010) I privilege the decline in militance over the decline in union membership or

coverage. This is because collective bargaining is only secured by the threat of the

labor stoppage.

What is the relationship between strikes and freedom, directly stated? If a

person cannot abruptly cease working, in what sense is that person a free citizen? If

a person cannot abruptly cease working jointly and simultaneously with confederates,

then in what sense can the cessation of labor be instrumentally e↵ective toward

political or economic ends? These are more informal statements of the argument in

the conclusion of Lambert (2005). The right to strike is indispensable to citizenship

because work itself is inextricably bound with social standing and inclusion. As

Judith Shklar argues, ”the dignity of work and of personal achievement is [integral

to] American civic self-identification” (1991, 1). And, like Shklar, I am persuaded

by the contention that citizenship is best understood as a kind of belonging. What

is more, I identify this right to belonging - the right to be a free and full part of

4By ’elite,’ I mean those citizens who are capable of causing serious trouble for politicians (i.e.damage their reelection chances) either unilaterally or in concert, given their number is small. So,then, labor leaders are only elites within the Democratic Party, largely as a result of the geographicconcentration of labor density in the United States (Dark 2000; Roof 2011).

181

a society - is in tension with the right to liberty as construed in the tradition of

American political thought. As Shklar points out (Ibid, 67), ”this vision of economic

independence, of self-directed ’earning’ as the ethical basis of democratic citizenship”

has been predominant since the advent of Jacksonian democracy; and, therefore, ”we

are citizens only if we earn” (Ibid). Indeed, it is not accidental that the Philadelphia

convention never considered extending the franchise to low-wage workers for it was

their considered view that their status of dependence on the pay packet disqualified

them from citizenship (Forbath 1999, 18-20).

In contrast to Lambert’s argument, upon which I am indeed reliant, I do not

defend the right to strike against the argument that the strikes are to be deplored for

their inconvenience (cf. Bok and Dunlop 1970, 229) - a core goal of the Taft-Hartley

amendments as presented in its statement of policy - in contrast, I take seriously the

tension that was and continues to be powerfully exploited by conservative politicians.

Thus, I am defending the right to strike against liberal and libertarian objections.

The result of the New Deal, with respect to the emergent policy of strike restriction,

was not a capitalist auto-correction; it reflected a majoritarian preference for anti-

democratic policy.

The dominant school of thought in the formation of the Roosevelt labor rela-

tions policy was the industrial pluralist school. These thinkers held that collective

bargaining was itself beneficial for society but only to the extent that it did not in-

volve protracted labor stoppages. This is because in the pluralist tradition held that

labor disputes were not born from class antagonism but rather from social complex-

ity in large enterprises. A leading pluralist of the 1930s, Arthur Ross, acknowledged

class antagonism thusly, ”One of the virtues of collective bargaining is that it permits

the formulation of limited issues which are amenable to resolution and blurs over the

182

large di↵erences of principle which can never really be settled” (1954, 532). More

pointedly, A.H. Raskin, a labor reporter and editor for the New York Times, argued

”I see no reason why in this institution alone [the labor relations order], of all facets

of our society, we should exalt the right to make war as the hallmark of industrial

civilization when we seek to exorcize it everywhere else” (1967, 136).

Conservative opponents to industrial unionism who proceeded on principle

would likely endorse Hayek’s position that ”the whole basis of our free society is

gravely threatened by the powers arrogated by the unions” (1960, 269). Hayek, and

the libertarians generally, were unalterably opposed to strikes because they involved

denying to other citizens the liberty of contract to replace the strikers. The e�cacy

of a strike is indeed grounded in coercion. Therefore, for my position to obtain, I

must make an appeal to a theory of justice; otherwise, I will merely be arguing for

the privileging of workers interests over those of capitalists (or consumers).5 It would

be tempting to appeal to Rawls’ original position here because any rational person

behind the veil of ignorance would prefer that the rights of strikers trump the right

to private property. This is because any individual would be far more likely to be

in the position of worker than that of capitalist. However, the Rawlsian concept is

based on individual liberty.

Thus, I focus on the right to strike as being a component of citizenship, an

essential property of which is the right to belonging. Shklar, in her seminal work

does not directly treat the right to strike but the work of TH Marshall upon whom

she draws does. ”Strikes usually involve breach of contract or the repudiation of

agreements. Appeal is made to some allegedly higher principle - in reality, though

this may not be expressly asserted, to the status rights of industrial citizenship”

5This is the legal (logical) positivist reading of my claim.

183

(Marshall and Bottomore 1950, 26). And this is crucial: a one man strike is just

a quitter. By entangling the group right claim to citizenship, the right to strike is

placed on a secure theoretical footing. If a group of workers cannot exercise the right

to strike, then they must be subordinate citizens; therefore, not truly citizens at all.

Note, this is distinct from saying that the worker is servile because of dependence on

the pay packet. I find this defense of the right to strike highly persuasive because

denying a citizenship right constitutes a civil excommunication.

Implicitly, I have been conceiving of rights as a claim that circumscribes the

legitimate grounds under which legal coercion might be applied. Necessarily, then,

equality of rights must infringe on liberty - broadly construed. Here, I propose,

following Dworkin (1977), that liberty be understood as a bundle of distinct rights

and not as a broad shield. In the Hayekian concept of liberty and commutative justice,

the right to several property is a shield against inequality. Yet, in contrast to Hayek’s

supposition that social justice is a fig leaf for politically motivated expropriation, I

emphasize that it is a purely theoretic contest between rights, that of citizenship and

property. I can further appeal to Michael Sandel (1996) who argues persuasively that

all rights are in some sense collective because all people must be embedded in a social

context of some kind. By definition, rights involve a claim and a corresponding duty

so they cannot exist in theoretical isolation.

Therefore, the Labor-Management Relations Act is a law that denies the social

dimension of humanity. The right to strike necessarily implies a protection from

job loss or criminal sanction as a result of industrial stoppage. In denying such

protection(s), the extant labor law serves to excise social citizenship from worker

citizenship. In their haste to delegitimize worker protest, the Congress promised to

contribute to the atomizing of society.

184

Though the rationale of maintaining the free flow of commerce can be found in

the Wagner Act, the Taft-Hartley amendments evidently turn the purpose of Wagner

labor law on its head. In this sense, though, it could be viewed as an anteceding case-

in-point for Schiller’s (2015) recent argument that the contradictions of progressive

policies adopted under the Democratic regime doomed postwar liberalism from the

outset. He summarizes thusly, ”When liberalism emerged from World War II, the

focus of workplace regulation was on ensuring the right of workers to organize labor

unions, and on forcing employers to bargain with those unions. By the middle of the

1970s, the regulation of employers had shifted. Paramount now was the protection

of the rights of individuals not to be discriminated against because of the immutable

characteristics” (Ibid, 240).

Another way of understanding the policy responses to labor militance in state

and nation is to note what the retrenchment chamber majorities did not do. To wit,

none proposed unconstitutional legislation to disband unions6, nor did they attempt

to disband unions on an intra-state basis. Indeed, as Paster (2015, 21) points out, the

politics surrounding the New Deal reforms suggest some strategic accommodation on

the part of capital.

Citizens who themselves had little stake in capitalist accumulation undeniably

had a stake in the delivery of consumer goods. Observers, such as Cohen (2003) have

made the case that in the social arena, that citizens tended to place a primacy on

their identity and interest as consumers. Halting production therefore pits workers

against consumers as well as capitalists. So, it is hard to argue that there was an

oligarchic distortion during the formation of New Deal policy.

6Such a measure would contravene the Wagner Act and so would have run aground upon theSupremacy clause.

185

Nonetheless, many observers of present-day politics would argue it tends toward

oligarchy (Bartels 2008; Gilens 2012; Gilens and Page 2014; Page and Winters 2015,

for example) and that the weakness of unions is a major cause of the United States’

distinctively unequal society (Morris and Western 1999; Stitglitz 2015, for example).

The weakness of the unions as a class vehicle stem chiefly not from membership

or density but from the absence of direct resistance. Indeed, the present-day gains

for the labor movement, such as they are, are dependent on extra-union class-based

social movement organizing (Fantasia and Voss 2004; Milkman and Voss 2004; Francia

2006).

Yet, worker resistance is essential to balancing capital. And this means more

than union organization, membership, or density - for example, the German experi-

ence which shows that regulatory liberalization and economic inequality have risen

even with high union density; this is because the unions prefer(red) insiders (cf. The-

len 2014; Collier 2015).

The present political economy has been increasingly characterized by non-

production work in recent decades and that trend should continue. If, in this new

economy, a CIO for the twenty-first century were to be formed, it would be based in

the service sectors and the health care sectors. While the nurses and janitors have

a worksite to seize, militance that denies medical care is as unsympathetic as strikes

that cut milk delivery or strikes that undermine a war e↵ort. Moreover, many of the

present-day workers and economic losers are in atomized workplaces, making organi-

zation di�cult - and making organization in the absence of militance useless. But,

again, I do not recommend the importance of unions qua unions, for they are only po-

litically meaningful for a broad swath to the extent that they engage in class politics.

Indeed, public sector unions are all too happy to be adjutant to the Democratic Party,

186

as Democratic majorities serve as a proxy for worksite co-determination.7

It should be noted that the use of replacement workers to break strikes has been

an essential tool in the capitalist countero↵ensive. The replacement worker doctrine

comes out of late 1930s Supreme Court cases (cf. Klare 1978; Barenberg 1993).

While the decline of the e�cacy strikes would be impossible without the employment

of replacement workers, the climate in which owners felt free to negate the right to

strike is an outgrowth of the context in which they were embedded; namely a context

in which the right to strike was narrowly legitimized.

Schiller (2015, 245-8) rightly argues that the collapse of the New Deal coali-

tion had its origins in the very formation of the policy itself; because the New Deal

o↵ered so little to working-class whites, those voters were quick to bolt for the Re-

publican Party. And because the labor relations framework made it di�cult for civil

rights demanders to believe that the union o�cers elected by white majorities would

allow the unions to petition and be vehicles for racial justice, there was little on of-

fer to Blacks. So, then, legal challenges raised by the civil rights demanders came

into direct conflict with union rights. Had there been a robust right to strike, the

industrial workforce would have been impelled to adopt racially harmonious policies

(like that of unions in the young CIO, the autoworkers, mine workers, mill workers,

rubber workers and textile workers), lest the Blacks be rational strikebreakers. So,

then, if workers could strike freely, the tension between majoritarian labor law and

minority-shielding employment law likely would not have made rivals of the merged

AFL and CIO and the civil rights movement. Also, as Lee (2014, 236) shrewdly

points out, the conservative movement, in its linking of union trouble with Black and

7Co-determination also implicates political theories of dignity which are important but distinctfrom the focus of this chapter.

187

women claimants helped blunt the white supremacist undertones of the the ascendant

conservative coalition.

The right to strike conceived as part and parcel of citizenship rights would lead

to the imposition of the worker’s voice into the production process and enshrine the

workers’ cartels as a civic institution. Industrial unions, then, were not merely a

vehicle for wage gains, they were a potential site for self-government and, therefore,

would have concomitant responsibilities to avoid the use of strike leverage as a means

to secure sub-group advantage. That is, these hypothetical, non-Taft-Hartley-bound

organizations, would have to elude the iron law of oligarchy. Whether this possibility

is theoretically possible is beyond the scope of this chapter, though I would share a

skeptical reader’s doubts. Nonetheless, this consideration is altogether distinct from

the question of citizenship as belonging or whether rights to full belonging in society

trump rights to property or liberty, broadly construed. It is worth pointing out, as

Lambert (2005, 194-6) does, that the thirteenth amendment can be credibly read

to guarantee the right to strike in its forbidding involuntary servitude; however, the

Supreme Court has summarily dismissed this idea without seriously engaging it. That

said, the ’switch in time’ involved a shift in doctrine as great as this one.

In summary, the right to strike is a group right that conflicts with individual

liberties. Because this group right is essential to citizenship (or, social belonging),

it is presented as prior to individual rights to property and contract in principle.

This argument generates the paradoxical insight that majoritarian preference ran

counter to democracy. And, indeed, this preference was borne out, for democracy

lessened and oligarchy flourished in the subsequent decades. This elides discussion

of the intervening political shocks of the civil rights movement and economic shocks

associated with deindustrialization, though; and this is why the present chapter has

188

taken normative theory as its focus, because the argument from principle is clearer

than the argument from consequence, as unhappy as those have been.8

1930s Backlash and the Present

The rollbacks that unions in Wisconsin faced in 2011 was one that overturned

an old law, in contrast to the backlash of the 1930s that immediately pared down a

new law. Upon winning the statehouse in President Obama’s first midterm, Governor

Scott Walker and his co-partisans set about applying a right-to-work law to state and

municipal employees (Stein and Marley 2013, 12-5).9

Also importantly, the backlash against the public sector unions was one made

in a context where the private sector unions had already been substantially weakened.

And, indeed, an important element of the Republicans’ advocacy for the rollback of

collective bargaining rights was rhetoric concerning fairness. If private sector workers

are exposed to the vicissitudes of the market, why should public sector workers be

insulated (Devinatz 2015, 9; Chesters 2016, 3)? But the previous decline of private

sector unions was not only rhetorical in its e↵ect. The weakness of existing unions also

left the public sector unions politically isolated. Though Republicans were variously

in support of public sector union security laws in the middle of the twentieth century,

Republican unified government is necessary for substantial rollbacks at present. The

partisan cleavage that emerged placed the public sector unions in an adjutant role

to the Democrats; but unlike the 1930 and 40s, this role was one that was happily

8Taking us back to our discussion of Smith’s (2000, Chapter 9) agnosticism over whether peoplewould be better o↵ under a social democratic regime. Here, our normative argument reveals thatthe middle tercile whose fortunes may or may not be better than their European counterpartshave a stunted citizenship, even if the inequality that arose from working-class enervation is notdemonstrably catastrophic for them.

9The Democrats, while in the majority, had a chance to enshrine the collective bargaining agree-ments into statute, but were unable to muster su�cient party unity. Two Democrats defected andthe measure failed in the upper chamber.

189

played by public sector workers.

Because of these limitations, a comparison between these eras is unlikely to

be deeply illuminating. However, the argument that union weakness arising from

the retrenchments against Wagner is connected to latter-day reversals is plausible.

Notably, victory of the ’Fight for 15’ in New York in 2015 was itself reliant on the ’baby

Wagner’ law adopted in 1937. In New York, workers picketed fast food restaurants

and dirsupted tra�c on throughfares while protesting for a statutory increase in the

minimum wage to fifteen dollars an hour. The New York Governor Andrew Cuomo

established an industry commission with the purpose of devising regulations toward

that end. These industrial commissions are authorized by the 1937 State Labor

Relations Act.

Altogether, this exercise in historical social science is of less utility to students

interested in union revival than it is to students of faded potentialities. This disser-

tation shows more about why and how the paths to American social democracy were

occluded at the height of labor’s power. The 1930s were the only time in US history

when business opposition to union formation and protest was overcome by the state

or the unions. But, even given this, the present-day condition of legal and political

weakness could not have been forestalled.

Limitations and Final Thoughts

This dissertation set out to see what could be gained from a revisiting of the

Wagner to Taft-Hartley interlude with a focus on the uneven experience of the states.

What came out of that analysis chiefly was an emphasis on the social control impetus

aroused by union militance among the public. This causal emphasis stands in contrast

to the emphasis that would be gleaned from study limited to the federal level where

190

scholarship has pointed to the preferences of the white supremacist delegation or the

preferences of corporate liberals within the Roosevelt administration.

In exploring this pattern of retrenching laws against strikes, we turned to the

media and opinion context in which the electorates were embedded. This investigation

was inconclusive in terms of determining whether the public was manipulated but

it did illuminate the hostility which worker protest encountered and, importantly,

suggested that majority preference had been served by retrenchment. Investigation

of these factors in the states highlighted the unease with which a muscular organized

labor movement was integrated into the existing partisan coalitional alignments. But,

just as importantly, the pattern of popular support for strike restraint carried with it

thorny normative implications, which formed the focus of the present chapter.

Crucial among the contributions of this dissertation is to locate the turning

point in US labor law history in 1938, as opposed to 1941, 1947, 1959, 1968, or 1980.

Before the Congress adjourned in 1939, Senator Claude Pepper (D-FL) rose to in-

tervene, ”I am unwilling to let this session of Congress end without lifting my voice

to decry the unrighteous partnership of those who have been willing to scuttle the

American government and the American people and jeopardize the peace of the world

because they hate Roosevelt and what Roosevelt stands for....I accuse that design-

ing alliance of a deliberate attempt to sabotage the first real e↵ort ever made in this

nation to secure for the workers of America industrial democracy and economic eman-

cipation. I accuse them of having prostituted their power to serve the US Chamber of

Commerce, the Manufacturer’s Association, and the beneficiaries of special privilege

who hate in their hearts the man who has tried to lighten the burden of toil on the

back of labor (quoted in Patterson 1967, 326). Here Senator Pepper has identified

the backlash as having occurred well before the traditional understanding. In short,

191

the limits of labor liberalism were reached at the height of the New Deal.

This is important because it strongly suggests that the present political order

is an altogether unhelpful basis on which to proceed for the labor-left. If attempts

to forge a new political economy at the apogee of militance were doomed, then the

present circumstance is untenable.

At the outset, I argued that the investigation of the states would also yield

counter-factual insight into the question of whether the Taft-Hartley Act might have

been forestalled or prevented in addition to facilitating testing theories of New Deal

policy change. In order to more convincingly make this case, further research might

structure the historical path as a sequence of variations represented by nodes that

connote the various strategies taken by state-level and local union a�liates during

this period.

The experience of the states bears upon union revival in that imagining a mus-

cular labor movement that is not subordinate to the Democratic Party requires imag-

ining a political order in which the right to strike is respected. In the absence of this

ability to exercise such power, it is nigh-impossible to suggest a di↵erent outcome.

Though unions may coalesce with social movement organizations (e.g. ’Fight for 15,’

the mass protest movement to raise the minimum wage to fifteen dollars per hour), the

success of that struggle is dependent on attention-grabbing disruptive protest which

raises the salience of their concerns. And, importantly, the major recent success that

protest movement secured in New York State was an exercise of gubernatorial author-

ity resting upon the old Doyle-Neustein bill, that is, the baby Wagner Act. But, here,

this should not be taken to undermine the normative argument presented above, for

the strike is a commercial (and, by extension, political) tool as well as an expression

of worker citizenship and personhood.

192

This dissertation is in some sense limited by its temporal scope. If we expanded

the time frame forward, we would contend with the role of public sector strikes in the

latter-day relationship between unionists and the Democrats. While the public sector

is strategically distinct from the private sector in terms of union purpose, protest,

and organization, the central problem of how everyday citizens wrest control from

the strong is the same. There is a cost to limiting the temporal scope to the adoption

of Taft-Hartley but the limitation of the right to strike in the private sector was an

essential component of the politics surrounding the incorporation of public sector

unions into state politics in the mid-20th century. Put another way, it is di�cult

to believe that the depredations su↵ered by latter-day public sector unions are not

a direct result of the losses incurred by unionists during the New Deal. Indeed, the

rhetoric of proponents of state laws that erode public union security justify the public

sector depredations on the basis that it would be unfair to exclude the government

workers from the precarity faced by private-sector workers.

The relationship between capital and labor that prevailed in the industrial po-

litical economy is extinct; but, the relationship between the two forged during the

New Deal persists. The relationship between white voters and the Democratic Party

has changed; the unions that support the Democratic Party are now more likely to

be public sector workers. The temptation for public sector unions to work for class

gains10 or to use militance is diminished because the politicians constitute the senior

managers. But, again, the relationship between the unions and the Democrats was

forged during the New Deal and set the pattern for the next century.

10Moe (2011) or DiSalvo (2015) would deem this an understatement and make the case thatthe public sector unions constitute a cartel of insiders extracting gains from the public till at theexpense of the citizenry. I am agnostic regarding diagnoses of latter-day unions. Su�ce to say,we have considered the roles and anticipations of militant worker activists and we can say that thedecision to bureaucratize and become adjutant to the Democratic Party was freighted with tradeo↵s.

193

The focus on the industrial political economy seems to limit the relevance of

the dissertation to the present circumstance, for the nature of work has changed. The

’less hierarchical organization of the workplace in algorithmic jobs seems to suggest

that workers are managing the managers as much as the other way around. However,

there is an ineluctable disparity in power between owners and managers and workers,

even if the workers are autonomous in how they go about meeting production goals.

The social conventions of work are apparently less hierarchical than the industrial

plants of the 1930s and 40s, but this is a veneer over the insecurity of the workers

that are also increasingly atomized. In the final instance, the fortunes of workers

remain dependent on their capacity to form e↵ectual collective organizations.

Crucially, the judiciary has been on the outskirts of this dissertation while their

causal role is quite central: judicial hostility explains why Lewis was right to form the

renegade CIO, and judicial amenability explains the transformation of the Wagner

Act from parchment into a social and political reality. This is consistent with Orren’s

(1992) and Forbath’s (1991) ’judicial supremacy’ argument: in the development of

the labor relations order, the courts have served to set the rules of the game.

A virtue of this dissertation has been to ask whether the racial segmentation

of the labor force and, indeed, US society was an essential component of the politics

that established the extant labor law regime. The experience of the states suggests

that the US would have had a comparatively restricted labor movement, even if the

Jim Crow states had been in a separate polity.11 However, in leaving race o↵ to the

side, the present-day applicability of the foregoing insights is again limited.

These limitations notwithstanding, the foregoing analyses of the states and

11A comparison with the experiences of other English-speaking nations would also lend insightonto this matter.

194

regions strongly suggest that worker-protesters were in an untenable position even at

the apogee of their might, and the resultant legal regime is more or less what the

mass public wanted: unions should be legally recognized but their protest should

be muted. The resultant original contributions are twofold: first, to identify a race-

neutral explanation for the prevalence of union restriction; second, to outline pre-

War limits to liberalism in the states. The subsequent course of events throughout

the twentieth century substantiates the logic of placing our focus squarely on strike

restraint across the nation and upon the hollowing out of the concept of freedom that

arises from a truncated right to strike. These are consequential because they place the

current oligarchic US politics in a perspective that highlights the absence of a class-

counterweight precisely because worker protest has been diminished greatly.

195

References

Ahlquist, John S. and Margaret Levi. 2013. In the Interest of Others. Princeton:Princeton UP.

Albig, William. 1937. ”The Content of Radio Programs, 1925-1935,” Social Forces16: 338-59.

Amenta, Edwin, Chris Bonastia and Neal Caren. 2001. ”US Social Policy in Com-parative and Historical Perspective.” Annual Review of Sociology, 27: 213-34.

Amenta, Edwin and Neal Caren et al. 2009. ”All the Movements Fit to Print,”American Sociological Review, 74(4): 636-56.

American Presidency Project. 2014. http://www.presidency.ucsb.edu/showelection.php?year=1936

Anderson, James A. 1971. ”The alliance of broadcast stations and Newspapers:The problem of information control,” Journal of Broadcasting and Electronic Me-dia, 169(1): 51-64.

Archer, Robin. 2007. Why is There No Labor Party in the United States? Princeton:Princeton UP.

——–. 1983. Values and Assumptions in American Labor Law. Amherst: Uni-versity of Massachusetts Press.Atleson, James B. 1998. Labor and the Wartime State Urbana, IL: University ofIllinois Press.

Amenta, Edwin and Neal Caren et al. 2009. ”All the Movements Fit to Print,”American Sociological Review, 74(4): 636-56.

Backstrom, Charles H. 1956. The Progressive Party of Wisconsin, 1934-1946. Madi-son: Wisconsin UP.

Bafumi, Joseph, Robert S. Erikson and Christopher Wlezien. 2010. ”ForecastingHouse Seats From Generic Congressional Polls: the 2010 Midterm Election.” PS:Political Science and Politics, 43(4): 663-36.

Barenberg, Mark. 1993. ”The Political Economy of the Wagner Act,” Harvard LawReview, 1379-1496.

196

Bartels, Brandon. 2008. ”Beyond ’Fixed versus Random E↵ects:’ a Framework forImproving Substantive and Statistical Analysis of Panel, Time-series Cross-sectional,and Multilevel Data,” The Society for Political Methodology : 1-43.

Bartels, Larry M. ”Messages Received: The Political Impact of Media Exposure”American Political Science Review 87(2): 267-85.

Baum, Matthew A. and Samuel Kernell. 2001. ”Economic Class and Popular Supportfor Franklin Roosevelt in War and Peace.” Public Opinion Quarterly 65(2): 198-229.

Baumgartner, Frank and Bryan Jones. 1993. Agendas and Instability in Ameri-can Politics. Chicago: University of Chicago Press.

Baylor, Christopher A. 2013. ”First to the Party: The Group Origins of the PartisanTransformation of Civil Rights” Studies in American Political Development 27(2):111-41.

Bensel, Richard. 1983. Sectionalism and American Political Development. NewHaven: Yale UP.

Berinsky, Adam J. 2006. ”American Public Opinion in the 1930s and 1940s: TheAnalysis of Quota-Controlled Sample Survey Evidence.” Public Opinion Quarterly70(4): 499-529.

——– and Eric Schickler. 2011. ”The American Mass Public in the 1930s and 1940s.”Roper Center for Public Opinion Research, University of Connecticut.

Bernstein, Irving. 1970. Turbulent Years: A History of American Workers, 1933-1939. Boston: Houghton Mi✏in.

Bok, Derek C. and John T. Dunlop. 1970. Labor and the American Community.New York: Simon and Schuster.

Boudreaux, Donald J. and Thomas J. DiLorenzo. 1993. ”The Protectionist Roots ofAntitrust.” The Review of Austrian Economics 6(2): 81-96.

Bridges, Amy. 1997. Morning Glories: Municipal Reform in the Southwest. Prince-ton: Princeton UP.

Bruno, Robert. 2009. ”Evidence of Bias in the Chicago Tribune Coverage of Or-ganized Labor,” Labor Studies Journal, 34(3): 385-407.

197

Caldeira, Gregory. 1987.”Public Opinion and the US Supreme Court: FDR’s Court-Packing Plan,” American Political Science Review, 81(4): 1139-53.

Cannon, Brian Q. 2004. ” ’What a Power We Will Be in This Land’: The LDSChurch, the Church Security Program and the New Deal,” Journal of the West 43(2):66-75.

Cannon, Brian Q. and Jessie L. Embry. 2009. Utah in the Twentieth Century.Logan, UT: Utah State University Press.

Carreiro, Joshua L. 2005. ”Newspaper Coverage of the US Labor Movement: TheCase of Anti-Union Firings,” Labor Studies Journal, 30(3): 1-20.

Charnock, Emily. 2012. ”The Rise of ’Political Action’: Labor Unions and theDemocratic Party.” APSA 2012 Annual Meeting Paper.

Chen, Anthony S. 2007. ”The Party of Lincoln and the Politics of State Fair Em-ployment Practices Legislation in the North.” American Journal of Sociology 112(6):1713-74.

——–. 2009. The Fifth Freedom: Jobs, Politics, and Civil Rights in the UnitedStates, 1941-1972. Princeton: Princeton UP.

Chesters, Graeme. 2016. ”Social Unionism and the Framing of ’Fairness’ in theWisonsin Uprising,” Local Economy 1(1): 1-16.

Chong, Dennis and James Druckman. 2007. ”A Theory of Framing and OpinionFormation in Competitive Elite Environments.” Journal of Communication, 57:99-118.

Cohen, Lizabeth. 2003. A Consumer’s Republic. New York: Knopf.Cohen, Sanford. 1948. State Labor Legislation, 1937-1947. Athens, OH :Ohio StateUniversity Press .

Collier, Ruth Berins. 2015. ”The High-Tech Economy, Work, and Democracy,”IRLE Working Paper 114-5.

Compa, Lance. 2015. ”Careful What You Wish For: A Critical Appraisal of Propos-als to Rebuild the Labor Movement,” New Labor Forum, 5(2): 1-6.

Council of State Governments. 1937. Book of the States. Chicago: Council of StateGovernments.

198

——–. 1939. Book of the States. Chicago: Council of State Governments.

——–. 1941. Book of the States. Chicago: Council of State Governments.

——–. 1943. Book of the States. Chicago: Council of State Governments.

——–. 1945. Book of the States. Chicago: Council of State Governments.

——–. 1948. Book of the States. Chicago: Council of State Governments.

Dark, Taylor. 1999. The Unions and the Democrats. Ithaca, NY: Cornell UP.

Davin, Eric Leif. 2000. ”Blue Collar Democracy: Class War and Political Revo-lution in Western Pennsylvania, 1932-1937,” Pennsylvania History 67(2): 240-97.

Davis, Mike. 1986. Prisoners of the American Dream New York: Verso.

Darowski, Joseph F. 2009. ”The WPA Versus the Utah Church,” in Utah in theTwentieth Century.

Deseret News. 1936-47. https://news.google.com/newspapers?nid=Aul-kAQHnToC(accessed May 19, 2016).

Devinatz, Victor G. 2015. ”Right-to-Work Laws, the Southernization of US LaborRelations, and the Decline of the US Trade Union Movement,” Labor Studies Journal40(1): 1-22.

Doellgast, Virginia. 2012. Disintegrating Democracy at Work. Ithaca, NY: Cor-nell UP.

Domho↵, G. William and Michael J. Webber. 2011. Class and Power in the NewDeal. Palo Alto: Stanford University Press.

Dubofsky, Melvyn. 1994. The State and Labor in Modern America Chapel Hill:University of North Carolina Press.

Dworkin, Ronald. 1977. Taking Rights Seriously. Cambridge: Harvard UP.

Eisinger, Robert M. and Jeremy Brown. 1998. ”Polling as a Means Toward Presi-dential Autonomy: Emil Hurja, Hadley Cantril and the Roosevelt Administration.”International Journal of Public Opinion Research, 10(3): 237-56.

199

Emery, Edwin. 1972. The Press and America: An Interpretative History of theMass Media. Englewood Cli↵s, NJ: Prentice Hall.

Entman, Robert M. 1989. ”How the Media A↵ect What People Think,” Journalof Politics 51(2): 347-70.

Erikson, Robert S. and Kent L. Tedin. 1981. ”The 1928-1936 partisan realignment:The case for the conversion hypothesis.” American Political Science Review 75(4):951-962.

Estlund, Cynthia L. 2002. ”The Ossification of American Labor Law.” ColumbiaLaw Review : 1527-1612.

Fantasia, Rick and Kim Voss. 2004. Hard Work: Remaking the American LaborMovement. Berkeley: University of California Press.

Farber, Henry S. and Bruce Western. 2001. ”Accounting for the Decline of Unionsin the Private Sector, 1973 to 1998.” Journal of Labor Research 22(3): 459-485.

Farhang, Sean and Ira Katznelson. 2005. ”The Southern Imposition: Congress andLabor in the New Deal and Fair Deal.” Studies in American Political Development19(1):1-30.

Fearon, James. 1991. ”Counterfactuals and Hypothesis Testing in Political Science,”World Politics 43(1): 169-95.

Ferguson, Thomas. 1984. ”From Normalcy to New Deal: Industrial Structure, PartyCompetition, and American Policy in the Great Depression.” International Organi-zation 38(1): 41-94.

Fine, Sidney. 1969.Sit-Down: The General Motors Strike of 1936-7. Ann Arbor:University of Michigan Press.

Fink, Gary M. 1974. Labor’s Search for Political Order: The Political Behavior ofthe Missouri Labor Movement, 1890-1940. Columbia: University of Missouri Press.

Firth, David. 1993. ”Bias Reduction of Maximum Likelihood Estimates,” Biometrika80(1): 27-38.

Fischer, David Hackett. 2005. Liberty and Freedom. New York: Oxford UP.

200

Flynn, Francis J. 2000. ”No News is Good News,” Industrial Relations, 39(1): 139-60.

Fones-Wolf, Elizabeth A. 1994. Selling Free Enterprise: The Business Assault onLabor and Liberalism, 1945-60. Chicago: University of Illinois Press.

——–. 2000. ”Promoting a Labor Perspective in the American Mass Media: Unionsand Radio in the CIO era, 1936-56” Media, Culture and Society 22: 285-307.

Forbath, William. 1991. Law and the Shaping of the American Labor MovementCambridge: Cambridge UP.

——–. 1999. ”Caste, Class and Equal Citizenship,” Michigan Law Review 98(1):1-99.

Foster, James C. 1975. The Union Politic: The CIO Political Action Committee.Columbia: University of Missouri Press.

Francia, Peter L. 2006. The Future of Organized Labor. New York: Columbia UP.

Freeman, Richard B. and James Medo↵. 1984. What Do Unions Do? New York:Basic Books.

——–. 1997. ”Spurts in Union Growth: Defining Moments and Social Processes,”in The Defining Moment: The Great Depression and the New American Economy inthe Twentieth Century ed. Michael D. Bordo, Claudia Goldin, and Eugene N. White,(Washington DC: National Bureau of Economic Research, 1997), 15.

Frymer, Paul. 2008. Black and Blue. Princeton: Princeton UP.

——–. 2010.”Labor and American Politics.” Perspectives on Politics 8(2): 609-616.

Funkhouser, G. Ray. 1973. ”The Issues of the Sixties: An Exploratory Study inthe Dynamics of Public Opinion,” Public Opinion Quarterly 37(1): 62-75.

Galenson, Walter. 1960. The CIO Challenge to the AFL. Cambridge: Harvard UP.

Gall, Gilbert. 1989. ”CIO and Democratic Alliance,” Labor Studies Journal, 2(2).

Gamson, William and Andre Modigliani. 1987. ”The Changing Culture of A�r-mative Action.” Political Sociology, 3: 137-77.

Gans, Herbert R. 1979. Deciding What’s News. New York: Pantheon.

201

Gerstle, Gary. 1991. Working Class Americanism: The Politics of Labor in a TextileCity, 1941-60. Princeton: Princeton UP.

Getman, Julius G. 2010. Restoring the Power of Unions. New Haven: Yale UP.

Gilens, Martin. 1999. Why Americans Hate Welfare. Chicago: University of ChicagoPress.

——–. 2012. A✏uence and Influence. Princeton: Princeton UP.

——–. and Benjamin I. Page. 2014. ”Testing Theories of American Politics: Elites,Interest Groups, and Average Citizens,” Perspectives on Politics, 12(3): 564-81.

Goldfield, Michael. 1987. The Decline of Organized Labor in the United States.Chicago: University of Chicago Press.

——–. 1989. ”Worker Insurgency, Radical Organization, and New Deal Labor Legis-lation.” American Political Science Review. 83(4): 1257-82.

Gourevitch, Peter. 1986. Politics in Hard Times. Ithaca, NY: Cornell UniversityPress.

Gourevitch, Alex. 2016. ”Quitting Work But Not the Job: Liberty and the Right toStrike,” Perspectives on Politics 14(2): 307-23.

Greenstone, J. David. 1969. Labor in American Politics. Chicago: University ofChicago Press.

Gri�th, Barbara. 1988. The Crisis of American Labor: Operation Dixie and theDefeat of the CIO. Philadelphia: Temple UP.

Gross, James A. 1974. The Making of the National Labor Relations Board: A Studyin Economics, Politics, and the Law. Albany: State University of New York Press.

——–. 1981. The Reshaping of the National Labor Relations Board. Albany, NY:State University of New York Press.

Hacker, Jacob S. and Paul Pierson. 2002. ”Business Power and Social Policy: Em-ployers and the Formation of the American Welfare State.” Politics and Society 30(2):277-326.

202

Harris, Howell J. 1982. The Right to Manage. Madison: University of WisconsinPress.

Hartley, Fred. 1947. in Legislative History of the Labor Management Relations Act,Congressional Record–House.

Hartz, Louis. 1955. The Liberal Tradition in America. New York: Houghton Mi✏in.

Hattam, Victoria. 1990. ”Economic Visions and Political Strategies: American Laborand the State, 1865-1896, Studies in American Political Development, 4(1): 82-129.

Havelock, Eric A. 1983. ”Forty-Five Years Ago: The Oshawa Strike.” Labour/LeTravail, 11(1):119-24.

Hayek, Friedrich. 1960. The Constitution of Liberty. Chicago: Regnery.

Heinze, Georg and Michael Schemper. 2002. ”A Solution to the Problem of Sep-aration in Logistic Regression,” Statistics in Medicine 21(16): 2409-19.

Hibbs, Doubglas A. 1976. ”Industrial Conflict in Advanced Industrial Societies,”American Political Science Review, 70(4): 1033-58.

Hinton, Wayne K. 1975. The New Deal Years in Utah. Logan, UT: Utah StateUP.

——–. 1986. ”The Economics of Ambivalence: Utah’s Depression Experience,” UtahHistorical Quarterly, 54(3): 268-85.

Holter, Darryl. 1990. ”Labor Law and the Road to Taft-Hartley: Wisconsin’s ’LittleWagner Act’.” Labor Studies 15(2): 20-47.

——–. 2001. ”Sources of CIO Success: The New Deal in Milwaukee,” Labor His-tory 41(2): 199-226.

Honey, Michael K. 1993. Southern Labor and Black Civil Rights. Champaign: Uni-versity of Illinois Press.

Ives, Irving. 1940. The Story of a Legislative Committee. Albany: Williams Press.

Iyengar, Shanto and Donald Kinder. 1987. News That Matters. Chicago: Uni-versity of Chicago Press.

203

Jervis, Robert. 1997. System E↵ects. Princeton: Princeton UP.

Katznelson, Ira. 1981. City Trenches. New York: Pantheon Books.

——–, Kim Geiger and Daniel Kryder. 1993. ”Limiting Liberalism: The South-ern Veto in Congress 1933-1950.” Political Science Quarterly 108(2): 283-306.

——– and Zolberg, Aristide (eds). 1986. Working Class Formation: Nineteenth-Century Patterns in Western Europe and the United States. Princeton: PrincetonUP.

Kellstedt, Paul M. 2003. The Mass Media and the Dynamics of American RacialAttitudes. Cambridge: Cambridge UP.

Kennedy, David M. 1999. Freedom From Fear: The American People in Depres-sion and War, 1929-45. New York: Oxford UP.

Killingsworth, Charles C. 1948. State Labor Relations Acts: A Study of Public PolicyChicago, IL: University of Chicago Press.

Kim, Jin Hee. 2006. Labor Law and Labor Policy in New York State, 1920s-1930sSeoul: Seoul National University Press.

Klare, Karl. 1978. ”Judicial Deradicalization of the Wagner Act and the Originsof Modern Legal Consciousness,” Minnesota Law Review, 62: 265-339.

Kornhauser, Arthur, Robert Dubin and Arthur Ross (eds). 1954. Industrial Con-flict. New York: McGraw.

Lambert, Josiah Bartlett. 2005. The Right to Strike and American Political De-velopment. Ithaca: Cornell UP.

Lee, R. Alton. 1974. Truman and Taft-Hartley. Columbia, MO: University of Mis-souri Press.

Lee, Sophia Z. 2014. The Workplace Constitution from the New Deal to the NewRight. Cambridge: Cambridge UP.

Leiserson, William M. 1937. ”Right and Wrong in Labor Relations.” Notes in Leis-erson Papers Box 21, Labor Legislation Folder, Wisconsin State Historical Society.

Lewis, David. 1973. Counterfactuals. Cambridge: Harvard UP.

204

Lichtenstein, Nelson. 1985. The Most Dangerous Man in Detroit. New York: BasicBooks.

——–. 1997. ”Taft-Hartley: A Slave Labor Law?” Catholic University Law Re-view 47: 763-91.

——–. 1998. ”Roll the Union On,” Institute for Public A↵airs : 18-23.

——–. 2002. ”Politicized Unions and the New Deal Model: Labor, Business, andTaft-Hartley,” in The New Deal and the Triumph of Liberalism Sidney M Milkis andJerome Mileur eds. Amherst: University of Massachusetts Press.

——–. 2003. Labor’s War at Home: The CIO in World War II. Philadelphia: Tem-ple University Press.

Lipset, Seymour Martin. 1989. Continental Divide: The Values and Institutionsof the United States and Canada. New York: Routledge.

Marshall, TH and Tom Bottomore. 1950. Citizenship and Social Class. London:Pluto Press.

Martin, Christopher R. 2004. Framed: Labor and the Corporate Media. Ithaca:Cornell University Press.

May, Dean L. 1987. Utah: A People’s History. Salt Lake City: University of UtahPress.

McCormick, John S. and John R. Silito. 2011. A History of Utah Radicalism Logan,UT: Utah State University Press.

Milkman, Ruth. 2006. LA Story: Immigrant Workers and the Future of the USLabor Movement. New York: Russell Sage.

——– and Kim Voss. 2004. Rebuilding Labor: Organizers and Organizing in theNew Labor Movement. Ithaca, NY: Cornell UP.

Millis, Harry A. and Emily Clark Brown. 1950. From the Wagner Act to Taft-Hartley: A Study of National Labor Policy and Labor Relations. Chicago: Universityof Chicago Press.

Mills, C. Wright. 1948. The New Men of Power. New York: Harcourt, Brace.

205

Milton, David. 1982. The Politics of US Labor. New York: Monthly Review Press,1982.

Milwaukee Journal. 1936-47. https://news.google.com/newspapers?nid=jvrRlaHg2sAC (accessed May 19, 2016).

Moody, Kim. 1988. An Injury to All. New York: Verso.Morris, Martina and Bruce Western. 1999. ”Inequality in Earnings at the Close ofthe Twentieth Century,” Annual Review of Sociology 623-57.

Newman, Katherine S. and Elisabeth Jacobs. 2010. Who Cares? Princeton: Prince-ton UP.

New York State Department of Labor. 1931. Annual Report of the Industrial Com-missioner Albany: Department of Labor.

New York State Labor Relations Board. 1939. Annual Report of the New YorkState Labor Relations Board Albany: Williams Press.

New York Times. 1933-1939. ProQuest Digital Archive: http://search.proquest.com.ezproxy.cul.columbia.edu/publication/45545

O↵e, Claus. 1985. Disorganized Capitalism: Contemporary Transformations of Workand Politics. Cambridge: MIT Press.

Orren, Karen. 1995. ”The Primacy of Labor in American Constitutional Devel-opment.” American Political Science Review 89(2): 377-388.

——–. and Stephen Skowronek. 2004. The Search for American Political Devel-opment. New York: Cambridge UP.

Page, Benjamin I. and Je↵rey Winters. 2009. ”Oligarchy in the United States?”Perspectives on Politics, 7(4): 731-51.

Page, Benjamin I. and Robert Y. Shapiro. 1992. The Rational Public. Chicago:University of Chicago Press.

Page, Benjamin I., Robert Y. Shapiro and Glenn R. Dempsey. 1987. ”What MovesPublic Opinion,” American Political Science Review, 81(1): 23-44.

Panagopoulos, Costas and Peter L. Francia. 2008. ”The Polls–Trends: Labor Unions

206

in the United States,” Public Opinion Quarterly, 72(1): 134-59.

Parenti, Michael. 1986. Inventing Reality: The Politics of the Mass Media. NewYork: St. Martin’s.

Paster, Thomas. 2015. ”Bringing Power Back In,” Max Planck Institute Discus-sion Paper 15(3).

Patterson, James T. 1969. The New Deal and the States: Federalism in Transi-tion Princeton: Princeton UP.

Phillips-Fein, Kim. 2011. ”As Great and Issue and Slavery or Abolition: EconomicPopulism, the Conservative Movement, and the Right-to-Work Campaigns of 1958.”Journal of Policy History 23(4): 491-513.

Pierson, Paul. 1996. ”The New Politics of the Welfare State.” World Politics 48(2):143-79.

——–. 2000. ”Not Just What, but When: Timing and Sequence in Political Pro-cesses.” Studies in American Political Development, 14(1): 72-92.

Pitkin, Hanna. 1967. The Concept of Representation. Berkeley: University of Cali-fornia Press.

Piven, Frances and Richard Cloward. 1977. Poor People’s Movements New York:Random House.

Plotke, David. 1996. Building a Democratic Political Order. Cambridge: Cam-bridge UP.

Polenberg, Richard. 1968. Reorganizing Roosevelt’s Government. Cambridge: Har-vard UP.

Powell, Allan Kent. 1977. ”Mormon Influence on the Unionization of Miners,” Jour-nal of Mormon History 4(1): 91-100.

——–. 2013. Next Time We Strike. Boulder, CO: Colorado UP.

Prasad, Monica. 2012. The Land of Too Much. Cambridge: Harvard UP.

Prior, Markus. 2006. ”The Incumbent in the Living Room: The Rise of Televisionand the Incumbency Advantage in US House Elections,” Journal of Politics 68(3):

207

657-73.

Putnam, Hilary. 2002. The Collapse of the Fact/Value Dichotomy and Other Es-says. Cambridge: Harvard UP.

Raskin, AH and John T. Dunlop. 1967. ”Two Views of Collective Bargaining”in Challenges to Collective Bargaining. Englewood Cli↵s, NJ: Prentice-Hall.

Robertson, David Brian. 1989. ”The Bias of American Federalism: The Limits ofWelfare-State Development in the Progressive Era.” Journal of Policy History 1(3):261-91.

Rogers, Joel. 1990. ”Divide and Conquer: Further ’Reflections in the DistinctiveCharacter of American Labor Laws’.” Wisconsin Law Review 1:1-149.

Rondinone, Troy. 2010. The Great Industrial War. New Brunswick, NJ: RutgersUniversity Press.

Roof, Tracy. 2012. American Labor, Congress, and the Welfare State. Baltimore:Johns Hopkins UP.

Rosenfeld, Jake. 2014. What Unions No Longer Do. Cambridge: Harvard UP.

Rosswurm, Steven (ed.). 1992. The CIO’s Left-led Unions. Newark: Rutgers UP.

Rubin, Beth A. and Larry Gri�n and Michael Wallace. 1983. ”Provided Only thattheir Voice was Strong”: Insurgency and Organization of American Labor from NRAto Taft-Hartley,” Work and Occupations 10: 325-47.

Rubin, Beth A. 1986. ”Class Struggle American Style: Unions, Strikes, and Wages,”American Sociological Review 51: 618-33.

Sabl, Andrew. 2015. ”Two Cultures of Democratic Theory,” Perspectives on Pol-itics, 13(2): 345-66.

Sandel, Michael. 1996. Democracy’s Discontent. Cambridge: Harvard UP.

Schiller, Dan. 1981. Objectivity and the News: The Public and the Rise of Com-mercial Journalism. Phildelphia: Univerity of Pennsylvania Press.

Schiller, Reuel. 2015. Forging Rivals: Race, Class, Law, and the Collapse of PostwarLiberalism. Cambridge: Cambridge UP.

208

Schickler, Eric and Devin Caughey. 2011. ”Public Opinion, Organized Labor, andthe Limits of New Deal Liberalism, 1936-1945,” Studies in American Political Devel-opment 21(1): 1-28.

Schmidt, Diane. 1993. ”Public Opinion and Media Coverage of Labor Unions,”Journal of Labor Research 14(2): 151-64.

Sexton, Patricia Cayo. 1993. The War on Labor and the Left. Oxford: WestviewPress.

Shapiro, Robert Y. 1998. ”Public Opinion, Elites, and Democracy,” Critical Re-view12(4): 501-28.

——– and Yaeli Bloch-Elkon. 2008. ”Do the Facts Speak for Themselves?” Crit-ical Review 20(1) 115-39.

——-. 2015. ”Public Opinion,” in The Oxford Handbook of American Political De-velopment. New York: Oxford UP.

Shefter, Martin. 1983. ”Regional Receptivity to Reform: Legacy of the Progres-sive Era,” Political Science Quarterly. 98(3): 459-83.

Shklar, Judith. 1991. American Ciitzenship: The Quest for Inclusion. Cambridge:Harvard UP.

Sinclair, Barbara. 1978. ”The Policy Consequence of Party Realignment,” Amer-ican Journal of Political Science 22(1): 83-105.

Skocpol, Theda. 1979. Protecting Soldiers and Mothers. Cambridge: Harvard UP.

——– and Kenneth Finegold. 1982. ”State Capacity and Economic Interventionin the Early New Deal,” Political Science Quarterly 97(2): 255-78.

——– and Michael Goldfield. 1990. ”Explaining New Deal Labor Policy.” Ameri-can Political Science Review 84(4): 1297-1315.

Smith, Mark A. 2000. American Business and Political Power. Chicago: Univer-sity of Chicago Press.

Stein, Patrick J. and Jason Marley. 2013. More Than They Bargained For: ScottWalker, Unions, and the Fight for Wisconsin. Madison: University of Wisconsin

209

Press.

Stepan-Norris, Judith, and Maurice Zeitlin. 1995. ”Union Democracy, Radical Lead-ership, and the Hegemony of Capital,” American Sociological Review : 829-850.

Stinchcombe, Arthur L. 1978. Theoretical Models in Social History. New York:Academic Press.

Stone, Katherine. 1981. ”The Post-War Paradigm in American Labor Law,” YaleLaw Journal 90: 1509-80.

Sussman, Leila A. 1945. ”Labor in the Radio News,” Journalism Quarterly : 207-15.

Swenson, Peter A. 2004. ”Varieties of Capitalist Interests: Power, Institutions, andthe Regulatory Welfare State in the United States and Sweden,” Studies in AmericanPolitical Development (18)1: 1-29.

Taft, Robert H. 1948. in Legislative History of the Labor Management RelationsAct, 1947. ”Congressional Record, Senate – June 23, 1947.

———. Legislative History ”Congressional Record, Senate – April 25, 1947.

Teaford, Jon C. 2002. The Rise of the States: Evolution of American State Gov-ernment. Baltimore: Johns Hopkins UP.

Thelen, Kathleen. 2014. Varieties of Liberalization. New York: Cambridge UP.

Thompson, Edward P. 2001. The Essential EP Thompson. New York: The NewPress.

Tomlins, Christopher L. 1985. The State and the Unions: Labor Relations, Lawand the Organized Labor Movement in America, 1880-1960. Cambridge: CambridgeUP.

Troy, Leo. 1957. Distribution of Union Membership Among the States, 1939 and1953. Washington, DC: National Bureau of Economic Research.

US Bureau of the Census. 1938. Statistical Abstract of the United States. Wash-ington DC: Government Printing O�ce.

——–. 1940. Statistical Abstract of the United States. Washington DC: Govern-

210

ment Printing O�ce.

——–. 1943. Statistical Abstract of the United States. Washington DC: Govern-ment Printing O�ce.

——–. 1946. Statistical Abstract of the United States. Washington DC: Govern-ment Printing O�ce.

——–. 1948. Statistical Abstract of the United States. Washington DC: Govern-ment Printing O�ce.

Truman, David B. 1971. The Governmental Process. New York: Alfred A. Knopf.

Vallely, Richard M. 1989. Radicalism and the States. Chicago: University of ChicagoPress.

Verba, Sidney and Kay Lehman Schlozman. 1977. ”Unemployment, Class Con-sciousness, and Radical Politics: What Did Not Happen in the Thirties,” Journal ofPolitics 39(2): 291-323.

Vogel, David. 1978. ”Why Businessmen Distrust Their State: The Political Con-sciousness of American Corporate Executives.” British Journal of Political Science8(1): 45-78.

——–. 1983. ”The Power of Business in America: A Re-Appraisal.” British Journalof Political Science. 13(1): 19-43.

Weatherford, M. Stephen. 2014. ”The Eisenhower Tradition: Labor Policy in theNew Political Economy,” Studies in American Political Development, 28(2): 201-23.

Weber, Max. 1949. ”Objective Possibility and Adequate Causation in HistoricalExplanation,” in The Methodology of the Social Sciences (ed. Max Weber), NewYork: Free Press.

Weir, Margaret. 2005. ”States, Race, and the Decline of New Deal Liberalism,”Studies in American Political Development: 19(2): 157-72.

——–. 2013. ”Beyond the Plant Gates: Postwar Labor and the Organizations Sub-structure of Liberalism.” Paper presented at the Twentieth Century American HistoryWorkshop, Columbia University.

White, Graham J. 1979. FDR and the Press. Chicago: University of Chicago Press.

211

Wisconsin Labor Relations Board. 1937. Report of the Wisconsin Labor RelationsBoard Madison: Wisconsin Labor Relations Board.

State of Wisconsin. 1937. Blue Book. Madison: Democrat Printing Company.

Witte, Edwin. 1937. ”Notes on Industrial Unionsim,” Wisconsin State HistoricalSociety, Witte Papers, Box 116, Folder 2.

Witwer, David. 2004. Corruption and Reform in the Teamsters Union. Urbana-Champaign: University of Illinois Press.

——–. 2005. ”Westbrook Pegler and the Anti-union Movement,”Journal of Ameri-can History, 92(2): 527-52.

Woolley, John T. 2000. ”Using Media-based Data in Studies of Politics.” Ameri-can Journal of Political Science, 44(1): 157-61.

Zaller, John. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cam-bridge UP.

Zeitlin, Maurice and L. Frank Weyher. 2001. ”Black and White, Unite and Fight,”American Journal of Sociology, 107(2): 430-67.

Zieger, Robert H. 1988. John L. Lewis: Labor Leader. New York: Twayne.

——–. 1994. American Workers, American Unions, 1920-1985. Baltimore: JohnsHopkins UP.

——–. 1995. The CIO, 1935-1955. Chapel Hill: University of North Carolina Press.

Zieger, Robert H. 1994. American Workers, American Unions, 1920-1985. Balti-more: Johns Hopkins UP.

Zorn, Christopher. 2005. ”A Solution to Separation in Binary Response Models,”Political Analysis 13(2): 157-70.

212

AppendixChapter 2:

There are twenty newspapers across the nation from January 1936 until January1948 used in these searches: The New York Times, New York Herald Tribune, DailyBoston Globe, The Washington Post, Los Angeles Times, The Sun, The ChristianScience Monitor, Chicago Daily Tribune, The Atlanta Constitution, Wall Street Jour-nal, Atlanta Daily World, Afro-American, The Chicago Defender (National edition),Philadelphia Tribune, The Pittsburgh Courier, Cleveland Call and Post, and Los An-geles Sentinel.

Search StringsFigures 1-3: All indexed strike articles: (((all(strike) AND labor AND at.exact(”Front Page/Cover Story”)) AND ti(labor OR strike) AND pd(19360101-19480101);John L. Lewis articles: (all(john l. lewis) AND at.exact(”Front Page/Cover Story”))AND pd(19360101-19480101); Racketeering articles: (all(union) AND racket* ANDat.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101); CIO articles:(all(CIO) OR all(C.I.O) OR all(C-I-O) AND at.exact(”Front Page/Cover Story”))AND pd(19360101-19480101); Communism articles: all(labor union) AND (com-munist* OR reds) AND at.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101)

Figures 4 and 5: John L. Lewis Articles: (all(john l. lewis) AND at.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101); Sidney Hillman articles:(all(sidney hillman) ANDat.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101); Philip Murrayarticles: (all(philip murray) AND labor union AND at.exact(”Front Page/CoverStory”)) AND pd(19360101-19480101); David Dubinsky articles: (all(david dubin-sky) AND at.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101); Harry Bridges articles:(all(harry bridges) AND at.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101); William Green articles: (all(William Green AND AFL) AND at.exact(”Front Page/Cover Story”)) AND pd(19360101-19480101)

Figure 11: Plea: all(labor union) AND (plea OR grievance) AND at.exact(”FrontPage/Cover Story” OR ”Article” OR ”Image/Photograph” OR ”Feature” OR ”Illustration” OR ”Front Matter” OR ”Editorial”OR ”Editorial Cartoon/Comic” OR ”News” OR ”Commentary”) AND pd(19350101-19490101); Demand: all(labor union) AND demand AND at.exact(”Front Page/CoverStory” OR ”Article” OR ”Image/Photograph” OR ”Feature” OR ”Illustration” OR”Front Matter” OR ”Editorial” OR ”Editorial Cartoon/Comic” OR ”News” OR”Commentary”) AND pd(19350101-19490101)

213

Negative FramingsFlint Sit-Down, January 1, 1937 to February 22, 1937:New York Times: ”sit-down strikers still defying an injunction to vacate;” ”Demandswhich are called conditions of settlement,” January 4, 1937, pg. 1; ”Alfred P. Sloan,president of General Motors, accepted the gage of battle o↵ered by the United Au-tomobile Workers of the Committed for Industrial Organizing,” January 5, 1937, pg.1; ”General Motors held out little hope tonight of an early settlement of the strikewhich has thrown more than 52,000 men out of work,” January 8, 1937, pg. 1; ”everye↵ort will be made to get men back to work and General Motors will never toler-ate the domination of its employees by a small minority” January 17, 1937, pg. 1;”employees are engaged in a back to work movement,” January 22, 1937, pg. 1; ”thePresident’s statement was a rebuke to John L. Lewis,” January 23, 1937, pg. 1; ”thestrike is an issue between property rights and no property rights,” January 24, 1937,pg. 1; ”employees being deprived of the right to work by a small minority who haveseized certain plants and are holding them as ransom,” January 28, 1937, pg. 1; ”Sloan applauded the demonstration of loyalty and described that over 100,000 havesignaled their desire to return to work,” ibid; ”these strikes have passed beyond thesit down stage into the occupational stage,” February 1, 1937, pg. 1; ”CongressmanReed of New York said that the sit-down strikes were the most un-American thing inAmerica,” February 5, 1937, pg 1;

New York Herald Tribune: ”The union has made collective bargaining impossible byusing coercion,” pg.1; ”The General Motors corporation today made public telegramssent to it by leaders of craft unions opposed to the United Auto Workers of theCongress for Industrial Organizing, January 9, 1937, pg. 1; ”The auto strike situationwas worse here in its potentiality for violence tonight than at any point since itsbeginning,” January 29, 1937, pg. 1; ”Sheri↵ Thomas Walcott mobilized the entirepolice force here to enforce an injunction order as Governor Frank Murphy ignoredan entreaty by the Sheri↵ to send troops,” February 4, 1937; ”John L. Lewis, the manbehind the strike which has paralyzed General Motors and the Governor,” February8, 1937, pg. 1.

Chicago Tribune: ”the UAW-CIO met to coordinate the movement they started andwhich has gone somewhat our of hand,” January 1, 1937, pg. 1; ”the manufacturersdo not want a showdown but Lewis is forcing one,” ibid; ”being led by a privategroup of labor dictators,” January 6, 1937, pg. 1; ”the strikes which are an unlawfuloccupation of General Motors plants,” January 10, 1937, pg. 1; ”the strikers mixedin riotous melees,” January 12,1937, pg. 1; ”The strike of General Motors that haskept 115,000 out of work and has cost workers and the corporation 10 million dollarsmoved toward a definite conclusion,” January 16, 1937, pg. 1; ”John L. Lewis saidwe helped Roosevelt lick industry, now let him help us lick industry,” January 22,1937, pg. 1; ”Lewis had baldly attempted to force upon the Roosevelt administrationthe obligation of repaying the political debt arising out of the support Lewis gave,”

214

January 23, 1937, pg. 1; ”General Motors corporation will continue its e↵orts Mondayto reopening of plants not closed by direct strikes in order to give work to as manyemployees as possible,” January 24, 1937, pg. 1; ”The violence flared out of controlwith the suddenness of a cyclone resulting in a new sit-down strike,” February 2,1937, pg. 1.

Baltimore Sun: ”Charging their enforced idleness to the dictates of a small groupof radical agitators,” January 22, 1937, pg. 1; ”C.I.O. is a group of agitators andracketeers,” ibid; ”General Motors says a back to work movement is supported by123,724 workers,” January 31, 1937, pg. 1; ”John L. Lewis strike dictator-general,”February 10, 1937, pg. 1.

Washington Post: ”’John L. Lewis strike dictator-general,” February 10, 1937, pg.1.

Christian Science Monitor: ”the seizure and possession of plants is not su�cient toclub an illegitimate demand out of employers when they have public opinion back ofthem,” February 12, 1937, pg. 1.

Boston Globe: ”John L. Lewis, generalissimo of the strike,” January 22, 1937, pg.1;

Los Angeles Times: ”Strike generalissimo, John L. Lewis, head of the Committee forIndustrial Organizing,” February 7, 1937, pg. 1.

Wall Street Journal: ”United Auto Worker chieftains drafted a program of far-reaching demands, threatening a company-wide walkout as an alternative to employercapitulation,” January 4, 1937, pg. 1; ”John L. Lewis knit more closely the divisionsof the army he is hurling at the mass production industries,” ibid; ”A rising tide ofresentment on the part of employees thrown out of work because of strikes engineeredby a minority was seen yesterday,” January 8, 1937, pg. 1; ”the corporation releasedfigures showing that 79 percent of employees have protested against strikes,” January22, 1937, pg. 1.

Lewis and Hillman, 1937-1940:New York Times: ”’Hillman has declined to go along with Mr. Lewis is attacking theNew Deal,” January 30, 1940, pg. 1; ”Lewis is Rebu↵ed on Defense Post,” August11, 1940, pg. 1; ”Hillman’s Union Will Stay in CIO, in Reply to Lewis Taunts,”November 21, 1940, pg. 1.

New York Herald Tribune: ”Hillman Fights in CIO for Labor Peace: Accepts LewisChallenge,” November 21, 1940, pg 1.

Los Angeles Times: ”CIO Warned Against Isms Menace to Labor by Hillman, Re-jecting the Suggestion of John L. Lewis” November 21, 1940, pg. 1.

Baltimore Sun: ”Hillman Vies Against Lewis For Contracts,” October 29, 1940, pg.1.

215

Wall Street Journal: ”New Dealers and members of Congress supporting new wageand labor standards have picked Sidney Hillman to appear before the joint committeemeeting and o↵set the unfavorable reaction that resulted from the testimony of JohnL. Lewis,” June 6, 1937, pg. 1.

Captive Mines, September 21, 1941 to November 23, 1941:New York Times ”President Roosevelt’s plea to Mr. Lewis to arrange for the con-tinued production in the captive mines was supplemented by a plea by Mr. Davis[National Defense Mediation Board Chair],” October 25, 1941, pg. 1; ”If a generalstrike were called the walkout would cripple the entire national defense program,”October 26, 1941, pg. 1; ”Steel men look upon the present strike as a culminationof the e↵ort of John L. Lewis to force a closed shop,” ibid; ”Mr. Roosevelt appealedtonight to the John L. Lewis of 1941 with the language of the Lewis of 1919, if Mr.Lewis refuses his patriotism will be impugned,” October 27, 1941, page 1; ”Come tothe aid of your country,” ibid; ”Sen. Byrd says there has been no more disgracefulor humiliating episode in American history,” October 28, 1941, pg. 1; ”On the heelsof John. L. Lewis’s refusal, the storm aroused in Congress beat higher today withsuggestions and plans for drastic anti-strike legislation for the defense industries,”Ibid; ”Senator Ellender declares John L. Lewis a traitor to our American ideals, anda menace to the peace and prosperity of our nation,” October 30, 1941, pg. 1; ”’Sen.Russell deems the walkout sabotage,” October 31, 1937, pg. 1; ”The president saidthe government will never compel the remaining to join the union by a governmentdecree. That would be too much like the Hitler methods toward labor,” November15, 1941, pg. 1; ”O�cials of the United Mine Workers of American, defying PresidentRoosevelt, late today ordered 53,000 men in the captive mines to withdraw,” Novem-ber 16, 1941, pg. 1; ”53,000 Told to Quit: UMW Heads Defy Roosevelt Warning,”November 16, 1941, pg. 1; ”I don’t mind strikes, but every minute? Why hell’s fire,I’m no rich man. I’m a working man says rank-and-file miner,” November 17, 1941,pg. 1; ”FDR described as not a valid point John L. Lewis’s principal gal for insistingupon acceptance of the union shop,” November 19, 1941, pg. 1; ”Faced with spread-ing sympathy stoppages of coal production, President Roosevelt was reported tonightto be nearing the point of cracking down on John L. Lewis,” November 21, 1941, pg.1; ”Carnegie Illinois says enough steel to build thirty destroyers was irrevocably lostto the national defense e↵ort,” November 23, 1941, pg. 1.

New York Herald Tribune: ”For the third time, your government through me asks youand the o�cers of the United Mine Workers to authorize an immediate resumptionof mining,” October 28, 1941, pg. 1; ”The president says he is considering legislationto strengthen the powers of the Executive to deal with labor controversies,” October29, 1941, pg. 1; ”The president has very strong cards in his hand with strong popularsupport for suppression of strikes which hold up defense production,” November 13,1941, pg. 1; ”Too much like Hitler methods,” November 15, 1941, pg. 1; ”Every [CIO]delegate is asking himself whether such a strike is justified in a time of national emer-

216

gency,” November 16, 1941, pg. 1; ”the president says Lewis strike justifications arenot valid,” November 19, 1941, pg. 1; ”Lewis Scorns Two Peace Plans of Roosevelt,”November 20, 1941, pg. 1.

Christian Science Monitor: ”Come now to the defense of your country, Presidentsays,” October 27, 1941, pg. 1; ”US to Run Mines Unless Lewis Yields; Nation BacksFDRs Fighting Talk,” October 28, pg. 1; ”The interruption of woe by strikes is thegreatest help the aggressors can get,” November 3, 1941, pg. 1.

Washington Post: ”Lewis Rejects Plea to End Mine Strike,” September 17, 1941, pg.1; ”in this crisis of our national life there must be uninterrupted production of coalfor making steel,” October 27, 1941, pg. 1; ”Anti-Labor Feeling Sweeps Capitol,”October 28, 1941, pg. 1; ”NAM president says if we can’t lick Lewis, then we betterlay o↵ Hitler,” November 19, 1941, pg. 1; ”President calls John L. Lewis’s reasonsnot valid,” ibid; ”Up to Mr. Lewis,” November 11, 1941, pg. 12; ”Three were shot instriker melee” November 22, 1941, pg. 1.

Baltimore Sun: ”President calls on John L. Lewis to come to the defense of yourcountry,” October 27, 1941, pg. 1; ”NAM Sounds Warning on Strike,” Ibid, 19;”Roosevelt makes fourth Plea to Lewis to End Strike,” October 30, 1941, pg. 1;”John L. Lewis ended his truce with the government and 53,000 of his miners wenton strike in defiance of the president’s edict,” November 16, 1941, pg. 1; ”We mustfind a way to halt these strikes in defense and we can no longer beg,” November 18,1941, pg. 1.

Los Angeles Times: ”uninterrupted production is essential to the preservation ofour freedoms, says President,” October 27, 1941, pg. 1; ”President Again Calls onLewis to Halt Strike,” October 28, 1941, pg. 1; ”public is against strikes in defenseindustries,” November 2, 1941, pg. 1; ”An e↵ective CIO strike has slowed the flow ofsteel and John L. Lewis belligerently defied the government to use troops,” November18, 1941, pg. 1; ”FDR Rebukes Lewis,” November 19, 1941, pg. 1; ”public hostilitytoward the unions together with repeated signs the administration is about ready toadvocate sharp labor control legislation were believed non factors in the mine workersdecision [to abort strike],” November 23, 1941, pg. 1.

The Atlanta Constitution: ”Roosevelt calls on Lewis to Avert Mine Strike Threat-ening US Defense,” October 25, 1941, pg. 1; ”uninterrupted production is essentialto the preservation of our freedoms, says President,” October 27, 1941, pg. 1; ”Ane↵ective CIO strike has slowed the flow of steel and John L. Lewis belligerently defiedthe government to use troops,” November 18, 1941, pg. 1; ”FDR Rebukes Lewis,”November 19, 1941, pg. 1.

Chicago Tribune: ”uninterrupted production is essential to the preservation of ourfreedoms, says President,” October 27, 1941, pg. 1; ”The president now finds himselfat the mercy of tyrants which he brought into being and established in power,”

217

October 29, 1941, pg. 1; ”Anti-strike was introduced to crush would-be dictators,”November 18, 1941, pg. 1; ’FDR Rebukes Lewis,” November 19, 1941, pg. 1; ”JohnL. Lewis bowed to the opposition of public opinion,” November 23, 1941, pg. 1.

Boston Globe: ”uninterrupted production is essential to the preservation of our free-doms, says President,” October 27, 1941, pg. 1; ”An e↵ective CIO strike has slowedthe flow of steel and John L. Lewis belligerently defied the government to use troops,”November 18, 1941, pg. 1; ”FDR Rebukes Lewis,” November 19, 1941, pg. 1; ”Gunfire fells 11 pickets in Pennsylvania,” November 22, 1941, pg. 1.

Wall Street Journal: ”Unions appear to hold the position that mediation is all rightwhen employers lose but it is all wrong when employers win,” November 11, 1941,pg. 1; Mr. Lewis clearly wants to maneuver the president into the role of an enemyof organized labor,” November 19, 1941, pg. 1.

Reconversion Strikes:New York Times: ”President Truman moved today to arrest a spreading wave ofindustrial strife, ordering a comprehensive reorganization of the labor functions ofgovernment,” September 18, 1945, pg. 1; ”James A. Krug, chairman of the War Pro-duction Board, expressed concern that strikes and work stoppages had deal industrialreconversion a severe blow and that they were the sole obstacle to the swift returnof long sought consumer goods,” October 6, 1945, pg. 1; ”City Faces Milk SupplyCut As Steel Strike O↵shoot,” January 22, 1946, pg. 1; ”Government Calls SoftCoal Strike National Disaster,” May 4, 1946, pg. 1; ”President Truman denouncedthe walkout today as nearing the status of a strike against the government,” May 9,1946, pg. 1.

New York Herald Tribune: ”President Truman acting to halt the strike wave which isendangering the nation’s reconversion program called on Congress to pass anti-strikerecommendations,” December 4, 1945, pg. 1.

Christian Science Monitor: ”President orders seizure of meatpacking plants, AFLwill work, CIO not known,” January 23, 1946, pg. 1.

Los Angeles Times: ”The Republican Party in recent months has gained groundsteadily with the public on the issue which voters themselves consider to be thenumber 1 question - regarding strikes and labor battles,” June 23, 1946, pg. 1.

Wall Street Journal: ”The CIO high command is pushing s sustained campaign fora planned American economy,” January 17, 1946, pg. 1.

Pre-Wagner:

Atlanta Constitution: ”The National Association of Manufacturers intends to proposelegislation this winter to ban general and sympathetic strikes. Association o�cialsannounced the plan tonight along with the opposition to the proposed Wagner labor

218

disputes bill...” January 15, 1935, pg. 1; ”A long-awaited report on wages in thecotton textile industry today brought from the United Textile Workers a number offresh demands and a definite threat of another bitter strike unless they are granted,”January 21, 1935, pg. 1.

219

Chapter 3:

Figures 1-5; 27: ”Are you in favor of labor unions?”

Figures 6-10: January 1937, February 1937 ”In the current General Motors strikeare your sympathies with the John L. Lewis group of striking employees, or withthe employers?” July 1937 ”John L. Lewis is trying to organize the workers of theFord Motor Company into a CIO (Congress of Industrial Organizations) labor union.Do you hope he succeeds?” December 1937 ”Would you like to see John L. Lewissucceed in organizing the steel industry?” May 1939 ”Which labor leader do you likethe best–William Green of the A.F. of L. (American Federation of Labor) or John L.Lewis of the C.I.O. (Congress of Industrial Organizations)?” November 1941 ”Whatis your opinion of John L. Lewis?” April 1942 ”John L. Lewis is planning to organizethe dairy farmers of the country into a branch of the C.I.O. (Congress of IndustrialOrganizations) Union. Do you favor or oppose, this movement to organize farmersinto unions?” June 1943 ”What is your opinion of John L. Lewis?”

Figures 11-15: April 1941, May 1941 ”Should the Government forbid strikes inindustries manufacturing materials for our national defense program, or should theworkers in those industries continue to have the right to go on strike?”; June 1941, July1941, August 1941, September 1941, October 1941, November 1941, December 1941”Should the government forbid strikes in defense industries, or should the workers inthose industries continue to have the right to go on strike?”

Figures 16-21: ”Do you think the Wagner Labor Act should be revised, repealedor left unchanged?”; ”In general, do you approve or disapprove of the RooseveltAdministration?”

Figures 22-26: February 1937 ”Do you favor President Roosevelt’s plan to increasethe size of the Supreme Court to make it liberal?” March 3, 1937 ”Are you in favorof President Roosevelt’s proposal to reorganize the Supreme Court?” March 12, 1937”Are you in favor of President (Franklin) Roosevelt’s proposal regarding the SupremeCourt?” April 1937 ”Are you in favor of President Roosevelt’s proposal to reorganizethe Supreme Court?” May 5, 1937 and May 12, 1937 ”Should Congress pass PresidentRoosevelt’s Supreme Court plan?” June 3, 1937 ”Should Congress pass the President’splan to enlarge the Supreme Court?”

220